HK1012477B

HK1012477B - Pcs pocket phone/microcell communication over-air protocol

Info

Publication number: HK1012477B
Application number: HK98113358.2A
Authority: HK
Inventors: Gary B. Anderson; Ryan N. Jensen; Bryan K. Petch; Peter O. Peterson
Original assignee: Xircom Wireless, Inc.
Priority date: 1994-03-21
Filing date: 1995-03-20
Publication date: 2006-06-09

Description

This invention relates to the field of communications, and particularly to communication systems using spread spectrum techniques and to over-the-air protocols for mobile telephones.

A mobile telephone system may generally comprise a set of "user stations", typically mobile and the endpoints of a communication path, and a set of "base stations", typically stationary and the intermediaries by which a communication path may be established or maintained. In a mobile telephone system, one important concern is the ability of mobile stations to communicate with base stations in a simple, flexible and rapid manner. The communication protocol between user stations and base stations should be rapid, so that user stations are not required to wait to establish a communication path. The protocol should be simple, so that user stations need not incorporate expensive equipment to implement it. The protocol should be flexible, so that user stations may establish communication paths in as many communication environments as reasonably possible.

Accordingly, it would be advantageous to provide a simple and flexible over-air protocol for use with a mobile telephone system. One class of systems in which this would be particularly advantageous is that of personal communication systems, particularly those with hand-held telephones in a microcell or other type of cellular communication system.

EP 0 328 100 describes a method for the handover of cells as a user station moves from the service zone of one base station to that of another.

The invention, which is defined in the appended claim, provides a method of establishing and maintaining a communication link between base and user stations.

The invention can be embodied in a simple and flexible over-air protocol for use with a mobile telephone system, such as a Personal Communication System (PCS) with hand-held telephones in a cellular communication system. The protocol can be adapted to "pocket phones", i.e., small hand-held telephones which may use a cellular communication technique, but the invention may be used with any cellular or mobile telephone system. The protocol defines a method in which user stations, such as cellular or mobile telephone handsets, communicate with one or more base stations to place and receive telephone calls. The protocol provides air-channel agility between base stations and user stations, while providing a secure voice or data link and the ability to handoff calls between base stations while they are in progress.

Each base station may have a set of "air channels" which it polls, e.g. by transmitting to each one in sequence. The air channels supported by each base station are referred to as a "polling loop" for a particular base station. A user station may receive information on an unoccupied air channel, receive the base station's transmission, and transmit information to the base station. Each base station may therefore simultaneously maintain communication with as many user stations as there are air channels in its polling loop. The ability of a user station to communicate on any unoccupied air channel makes the protocol air-channel agile. Each base station continually transmits on each one of its air channels in a predetermined sequence. Each base station transmission may be followed by a first gap, a user station transmission (if some user station attempts to communicate), and a second gap, before the base station transmits on the next air channel. A base station transmission, first gap, user station transmission, and second gap are collectively called a "minor frame". A polling loop in which each air channel is polled is called a "major frame".

Stability of user station and base station clocks may define the air channels, gaps, and minor frames. The user station may synchronize itself to the base station's clock by detecting a minor frame and by adjusting its clock to be in synchrony with the base station when the first bit sequence of the minor frame is detected. The stability of the user station and base station clocks may then hold the user station and base station in synchronization, as long as the user station is periodically able to receive transmissions from the base station. Should reception in either direction be interrupted for too long, the base station and user station clocks may drift apart and the user station may need to reacquire the transmission from the base station.

Handoffs are preferably initiated from the user station which continually monitors available air channels from the same and competing base stations during dead time. A user station may handoff within the same polling loop to establish communication in a new minor frame, or may handoff in such a manner to establish communication in a new minor frame within a polling loop of a different base station. In the latter case, a base station controller may assist in transferring the call from one base station to another.

Variable data rates provided in another aspect of the present invention. A user station may increase its data rate by transmitting and/or receiving in multiple minor frames during a major frame, or may reduce its data rate by transmitting and/or receiving in fewer than every major frame.

The invention will be described below with reference to the following drawings, in which:

Figure 1A is a diagram of a communication system having base stations and user stations;
Figure 1-1 is a diagram of a communication system with alternate network interconnections;
Figure 1-2 is a diagram of a network architecture showing various system components;
Figure 1-3 is a diagram of a network architecture showing connections between base stations and a network;
Figures 1-4, 1-5, 1-6 and 1-7 are diagrams of network architectures showing various system components;
Figures 1-8 and 1-9 are diagrams of a handset air channel acquisition procedure for an embodiment of the invention;
Figure 2 is a diagram of frame and message formats in a polling loop;
Figure 2-1 is a diagram of a preferred cellular environment in which the invention may operate;
Figure 2-2 is a diagram of a TDMA/TDD frame structure for an embodiment of the invention;
Figure 2-3 is a diagram of a polling loop for an embodiment of the invention;
Figure 2-4 is a block diagram of a speech coder for an embodiment of the invention;
Figure 2-5 is a functional block diagram of a speech coder control path for an embodiment of the invention;
Figure 2-6 is a diagram of a packet structure for an embodiment of the invention;
Figures 2-7 and 2-8 are diagrams showing time slot assignments for an embodiment of the invention;
Figure 2-9 is a diagram of a speech coder buffer structure for an embodiment of the invention;
Figure 3 is a diagram showing formats for message types; and
Figures 3-1 and 3-2 are estimated cell radii coverage charts for an embodiment of the invention.

It is contemplated that communication between base stations and user stations will be conducted using a spread-spectrum technique. There are at least three methods for establishing synchronization and communication, each preferably using an M-ary technique in which multiple bits of data are transmitted for each spread-spectrum symbol, e.g., by transmitting and receiving multiple different spreading codes, and interpreting the received one of those multiple different spreading codes at the receiver to indicate multiple data bits. Synchronization may be accomplished either by (1) automatic synchronization disclosed in US 5,761,239, US 5,790,591 and US 5,724,383, by (2) synchronizing with matched filters, by (3) demodulation and despreading using sliding correlators, or by (4) a combination of these techniques, e.g., matched filters for synchronization plus sliding correlators for demodulation and despreading, or matched filters for synchronization plus autosynchronization for demodulation and despreading.

Figure 1A is a diagram of a communication system having base stations and user stations.

A communication system 101 for communication among a plurality of user stations 102 may include a plurality of cells 103, each with a base station 104, typically located at the center of the cell 103. Each station (both the base stations 104 and the user stations 102) may generally comprise a receiver and a transmitter. The user stations 102 and base stations 104 preferably communicate using time division multiple access (TDMA) or time division duplex (TDD) techniques as further described herein, in which specified time segments or major frames are divided into assigned time slots or minor frames for individual communication.

Figure 2-1 is a diagram of a preferred cellular environment in which the invention may operate. A geographical region is divided into a plurality of cells 103. Associated with each cell 103 is an assigned frequency and an assigned spread spectrum code. Preferably, three different frequencies F1, F2 and F3 are assigned in such a manner that no two adjacent cells have the same assigned frequency F1, F2 or F3. The effect of such a frequency reuse pattern is to minimize interference between adjacent cells.

To further reduce the possibility of intercell interference, different orthogonal spread spectrum codes C1 through C6 are assigned as shown in adjacent clusters 110. Although six spread spectrum codes C1 through C6 are shown in Figure 2-1, it is contemplated that fewer or more spread spectrum codes may be suitable depending upon the particular information. Further information regarding a preferred cellular environment may be found in US 5,402,413, entitled "Three Cell Wireless Communication System" filed on April 8, 1991 in the name of Robert C. Dixon.

The use of spread spectrum for carrier modulation permits a very efficient frequency reuse factor of N=3 for allocating different carrier frequencies F1, F2 and F3 to adjacent cells 103. Interference between cells 103 using the same carrier frequency F1, F2 or F3 is reduced by the propagation loss due to the distance separating the cells 103 (no two cells 103 using the same frequency F1, F2 and F3 are less than two cells 103 in distance away from one another), and also by the spread spectrum processing gain of cells 103 using the same carrier frequencies F1, F2 or F3.

The preferred spread spectrum bandwidth may differ according to the frequency band of operation. When operating in the PCS A, B or C frequency bands, each of which is 15 Mhz wide, the center frequencies F1, F2 and F3 are preferably located at 2.5 Mhz, 7.5 Mhz, and 12.5 Mhz, respectively, from the lowest band edge of the A, B or C frequency band.

The PCS D, E or F bands, on the other hand, are each 5 Mhz wide, which is the same bandwidth as a preferred spreading bandwidth for a spread spectrum signal used in the particular cellular environment. Consequently, a single carrier frequency is placed in the center of the D, e or F band, and a frequency reuse factor of N=1 is used because the spread spectrum signal covers the entire available bandwidth. Because an N=1 frequency reuse pattern is used, the required intercell interference rejection must be obtained by spread spectrum code orthogonality and/or the use of sectorized antenna patterns. The exchange of interfering air channels or time slots, as described elsewhere herein, may also be used to mitigate intercell interference.

When operating in the PCS unlicensed band, which has a bandwidth of 20 Mhz divided into individual channel only 1.25 Mhz wide, the spread spectrum chipping rate may be reduced to approximately 1.25 Mcps. The TDMA burst rate, or number of TDMA time slots (or minor frames) in each polling loop, may also be reduced to maintain the required spread spectrum processing gain for rejecting intercell interference. A non-spread spectrum TDMA/TDD signal modulation format for operation in the unlicensed band may also be provided.

Figure 1-2 is a diagram of a network architecture showing various system components.

A preferred communication system is designed around an object-based software architecture which allows for flexibility in interconnection to various networks including public switched telephone networks, AIN, GSM and IS-41 network infrastructures. It is also contemplated that the communication system may interface with a cable television distribution network; however, such an interface may require the addition to the cable television network of a switch architecture, two-way amplifiers, redundancy, and, in order to use the coaxial portion of the cable TV network, a remote antenna subsystem to extend coverage from a base station 104.

The overall system thus provides flexibility to interface with a variety of different networks depending upon the desired application. To allow interconnection to diverse networks, the system uses internal communications based on ISDN messages, called "notes", for passing necessary information among components within the system. These "notes" are so named as not to confuse them with the ISDN specific protocol itself. Network messages (based on, e.g., Q.921, Q.931 protocols, or others) are converted by the system into "notes" for efficient operation within the hardware platform.

In Fig. 1-2 is shown various components of a preferred system architecture including a plurality of base stations 104 for communicating with user stations 102. Each base station 104 may be coupled to a base station controller 105 by any of a variety of linking means 109 including, for example, local area data access (LADA) lines, T1 or fractional T1 lines, ISDN BRI's, cable TV lines, fiber optic cable, digital radio, microwave links, or private lines. As an illustration shown in Fig. 1-2, a plurality of base stations 104 may be coupled to base station controller 105 by first connecting to a coaxial cable 111 which is thereafter coupled to a fiber optic cable 113 at a fiber node 112. The fiber optic cable 113 is coupled to the base station controller 105 as shown.

Each base station controller 105 may be connected to a network 106 such as a public switched telephone network (PSTN) or a personal communications system switching center (PCSC) by a variety of network links 108, which include the same basic categories of transport means as the linking means 109. Base station controllers 105 may also connect to the network 106 via an X.25 link 114.

The system of Fig. 1-2 also incorporates the use of "intelligent" base station (IBS) 107 compatible with LEC-based AIN architecture that may be connected directly to a network 106 without the interface of a base station controller 105. The intelligent base stations 107 may therefore bypass the base station controllers 105 for local handoffs and switching, and instead perform these functions via the network 106. In AIN based architectures, signaling between network elements may be carried out using standard signaling protocols including, for example, SS7 and IS-41.

In operation, the base stations 104 format and send digital information to the base station controller 105 (or directly to the network 106 in the case of an intelligent base station 107). The base station controllers 105 concentrate inputs from multiple base stations 104, assist handoffs between base stations 104, and convert and format channel information and signaling information for delivery to the network 106. The base station controllers 105 may also manage a local cache VLR database, and may support basic operations, administration and management functions such as billing, monitoring and testing. Each base station controller 105, under control of the network 106, may manage local registration and verification of its associated base stations 104 and may provide updates to the network 106 regarding the status of the base stations 104.

The network 106 connects to the base station controllers 105 for call delivery and outgoing calls. The connection between the network 106 and a base station controller 105 may utilize the Bellcore "Generic C" interface which includes Q.921, Q.931 and modifications to Q.931.

Intelligent base stations 107 may use ISDN messaging for registration, call delivery and handoff over a public telephone switch. The intelligent base station 107 may have all the general capabilities of a base station 104 but further incorporate a BRI card, additional intelligence and local vocoding. The connection between the network 106 and an intelligent base station 107 may utilize the Bellcore "Generic C" interface which includes Q.921, Q.931 and modifications to Q.931.

If the network 106 is a GSM network, then base stations 104 may connect to the network 106 through a defined "A" interface. Features and functionality of GSM are passed to and from the base stations 104 over the "A" interface in a manner that is transparent to the end user.

As noted, the system may also interconnect to cable television distribution networks. The base stations 104 may be miniaturized to the point where they can be installed inside standard cable TV amplifier boxes. Interfacing may be carried out using analog remote antenna systems and digital transport mechanisms. For example, T1 and FT1 digital multiplexer outputs from the cable TV network may be used for interfacing, and basic rate (BRI) ISDN links to transport digital channels.

Cell site diagnostics may be performed remotely through either the control channel on the digital link resident in the base station 104 or a dial up modem for some implementations. Such diagnostics may be performed on each component board of the base station 104. In addition, the base stations 104 and base station controllers 105 may be remotely monitored and downloaded with updated software as required. Similarly, user stations 102 can also be downloaded with software over air channels as required for maintenance purposes or for system upgrades.

The user stations 102 comprise in one embodiment mobile handsets capable of multi-band and/or multi-mode operation. The user stations 102 may be multi-mode in that they may be capable of either spread spectrum communication or conventional narrowband communication. The user stations 102 may be multi-band in the sense that they may be set to operate on a plurality of different frequencies, such as frequencies in either the licensed or unlicensed frequency bands.

For example, a user station 102 may be set to operate on any frequency between 1850 and 1990 MHz in 625 kHz steps. Thus, each user station 102 may have a frequency synthesizers which can be programmed to receive and transmit on any one of 223 frequencies. If the user station 102 operates solely in the licensed PCS band, however, the programmable frequency steps may be in 5 MHz increments, in which case the first channel may be centered at 1852.5 MHz, the next at 1857.5 MHz, and so on. If operating in the isochronous band between 1920 and 1930 MHz, the first channel may be centered at 1920.625 MHz, and the channel spacing may be 1.25 MHz across the remainder of the isochronous band. The user stations 102 need not operate in the 1910 to 1920 MHz band, which is reserved for asynchronous unlicensed devices.

Further detail regarding the multi-band and multi-mode aspects of user stations 102 may be found in US Patent 5,887,020, 5,815,525, 5,796,772, 5,790,587 and 5,694,414, as well as US Patent 6,389,059. The multi-band, multi-mode capability enables the user stations 102 to take advantage of variety of diverse system architectures as described herein, and to interface with various different networks with a minimum of hardware or software adjustments.

Base stations 104, like user stations 102, may also be provided with multi-band and multi-mode capabilities as described above.

FRAME AND MESSAGE FORMATS

Figure 2 shows frame and message formats in a polling loop.

In a single cell 103, a base station 104 may poll user stations 102 in the cell 103. The base station 104 may repeatedly transmit a major frame 201, comprising a sequence of minor frames 202. As noted herein, each minor frame 202 may comprise a polling exchange for a single user station 102, while each major frame 201 may comprise a complete polling sweep of user stations 102 in the cell 103.

The base station 104 may conduct its polling exchanges using a set of air channels 203. Each of the air channels 203 may comprise a separate transmission channel, such as a separate frequency band for FM or AM encoding, a separate spreading code for spread-spectrum encoding, a separate spatial location, or other division of communication slots between base stations 104 and user stations 102. The base station 104 may poll every one of its air channels 203 in a predetermined sequence in a single major frame 201.

While the base station 104 may poll every one of its air channels 203 in a single major frame 201, but it will be clear to those of ordinary skill in the art, after perusal of this application, that the base station 104 may restrict its poll to only a portion of its air channels 203 in each major frame 201, so long as all air channels 203 are eventually polled, and in an order so that each user station 102 may determine in which minor frame 202 it should respond.

Each minor frame 202 may comprise a base transmission 204 by the base station 104, a first gap 205, a user transmission 206 by a user station 102 (if any user station 102 responds), and a second gap 207. During the base transmission 204, a user station 102 desiring to establish a communication path may receive the base transmission 204 and determine if the air channel 203 is occupied or not. If not occupied, the user station 102 may respond with its user transmission 206.

In order to provide efficient service in low density rural areas, cell radii can be extended to large distances (e.g., beyond 8 miles) by providing the increased guard times as would be required for the longer round trip propagation delays encountered in the larger cells. Cells with large radii can be supported by reducing the number of minor frames 202 per major frame 201 to a lesser number (e.g., from 32 to 25). Since such large cell radii will ordinarily be deployed in low population density areas, reduced cell capacity caused by the smaller number of minor frames 202 per major frame 201 is not a severe drawback.

A base transmission 204 may comprise a header field 207, which may be a fixed length of sixteen bits, a D field 208, which may be a fixed length of eight bits, and a B field 209, which may be a fixed length of 160 bits, or may be a variable length. When using a variable-length B field 209, the variable length may be determined in response to the polling loop time and the data rate which must be supported. For example, in a 30-channel system, the B field 209 may be 160 bits long.

The user transmission 206 may comprise like fields as the base transmission 204.

The header field 207 may comprise an origin bit 210, which may be a "1" bit for base transmissions 204 and may be a "0" bit for user transmissions 206. Other parts of the header field 207 may indicate information about the base transmission 204 or user transmission 206 itself, e.g., what type of message the base transmission 204 or user transmission 206 comprises. The header field 207 may also comprise a CSC or CRC code 211 (a cyclic redundancy check) having four bits.

The D field 208 may comprise control information to be communicated between base stations 104 and user stations 102 once a communication link is established. This control information may generally be used for ISDN communication between base stations 104 and user stations 102, such as control information generally communicated using the ISDN "D channel". Because the D field 208 is separate from but simultaneous with the B field 209 which normally handles the bulk of information transfer due to its higher data rate, the D field 208 may be used for paging applications, notifications (e.g., voice mail), short message service (similar to GSM), or other user applications. Thus, the simultaneous nature of the D field 208 and the B field 209 allows messaging functions even when the user station 102 is "in use".

During link expansion, described with regard to figure 3 herein, the D field 208 may also comprise a user nickname 212 for communication from the base station 104 and a designated user station 102. The user nickname 212 may comprise a temporary identifier for the user station 102 selected by the base station 104.

The B field 209 may comprise data, voice (encoded digitally or otherwise), or other information. The B field 209 may also comprise specified information for establishing communication links between base stations 104 and user stations 102. The B field 209 may also comprise its own FCW or CRC code 211 having sixteen bits (with 160 bits of information, a total of 176 bits).

There may be 32 air channels 203; the major frame 201 may therefore comprise 32 minor frames 202 in sequence. Thus, each minor frame 202 may be about 307 microseconds long, each air channel 203 (in a TDD or TDMA system) may be about 667 microseconds long, and each major frame 201 may be about 20 milliseconds long. There may be 160 bits transmitted per air channel 203; thus the 32-channel system would have about a 256 kilobits/second total two-way data rate. Other time values are shown in the figure.

Information may be transmitted at a rate of five bits each 6.4 microseconds, using a 32-ary code-shift keying technique. Thus, each 6.4 microseconds, one of 32 different codes may be transmitted, with 32 different possibilities equalling five bits of information. As an alternative, one of 16 different codes may be transmitted, with an additional phase bit on the carrier (or, in a second alternative, more than one phase bit on the carrier), again with 32 different possibilities equalling five bits of information.

A minor frame 203 may operate in an asymmetric mode in the sense that the greater portion of a minor frame 202 is devoted to either the base transmission 204 or the user transmission 206. High speed data transport in either direction (i.e., from the base station 104 to the user station 102, or vice versa) can be provided in the asymmetric mode, with or without acknowledgment and/or ARQ.

A particular sub-mode of the above described asymmetric mode may be referred to as broadcast mode in which essentially the entire minor frame is devoted to one-way communication. In the broadcast mode, one or more broadcast sub-channels may be identified by a special broadcast identifier. Up to 255 broadcast channels may be so identified. For these point-to-multipoint applications, broadcast frames are not acknowledged.

Control Pulse

A user station 102 in a cellular environment preferably has means for controlling transmission power to avoid interference with adjacent cells. Unlike a fixed station environment, in which antenna locations, patterns and fixed station transmission power may be adjusted for minimal interference with other fixed stations, the nature of a cellular environment with mobile user stations 102 is such that there can arise conflict between user stations 102 at intersecting cell boundaries. This creates the need for some power control in the user stations 102. For example, a user station 102 operating at the boundary of coverage of a base station 104 may need to transmit at full power to stay in contact. On the other hand, a user station 102 operating relatively close to its own base station 104 may not need to transmit full power to have good contact. By proper power control, user stations 102 may maintain adequate contact with base stations 104 without unduly interfering with neighboring cell transmissions, allowing RF channel reuse in nearby cells. Power control may also reduce interference with fixed microwave users and conserve battery power in user stations 102 such as handheld units.

Power control is achieved by use of a power control pulse transmitted periodically from each user station 102. After establishment of a communication link, described with regard to Fig. 3 herein, a control pulse time 213 and a third gap 214 may be reserved just prior to the start of the minor frame 202, in which the user station 102 transmits a control pulse 215. The control pulse 215 provides to the base station 104 a power measurement of the air channel 203 indicative of the path transmission loss and link quality. Each user station 102 generally transmits its control pulse 215 in the minor frame 202 allocated to it (e.g., seized by the user station 102).

The control pulse 215 may be received by the base station 104 and used by the base station 104 to determine information about the communication link it has with the user station 102. For example, the base station 104 may determine, in response to the power, envelope, or phase of the control pulse 215, the direction or distance of the user station 104, and the degree of noise or multipath error to which the communication link with the user station 102 may be prone.

In response to receiving the control pulse 215, the base station 104 determines the quality of the received signal including, for example, the received power from the power control pulse 215 and the signal-to-noise or interference ratio. The base station 104 then sends a message to inform the user station 102 to adjust its power if needed. Based on the quality of the received signal, the base station 104 may command the user station 102 to change (increase or decrease) its transmit power by some discrete amount (e.g, in minimum steps of 3 dB) relative to its current setting, until the quality of the control pulse 215 received by the base station 104 is above an acceptable threshold.

Similarly, if the base station 104 knows the power setting of the user station 102, then the base station 104 can adjust its own power as well. The base station 104 may adjust its power separately for each minor frame 202.

A preferred power control command pulse from the base station 104 to the user station 102 may be encoded according to Table 5-1 below:

Table 5-1

Power Control Command	Adjustment
000	No change
001	-3 dB
010	-6 dB
011	-9 dB
100	+3 dB
101	+6 dB
110	+12 dB
111	+21 dB

Although preferred values are provided in Table 5-1, the number of power control command steps and the differential in power adjustment between steps may vary depending upon the particular application and the system specifications.

While power control is thus desirable, a problem in some conventional TDMA systems is that the length of the polling loop (e.g, the major frame 201) is too long to allow the latest user transmission to be very useful for estimating the channel losses and impairments. In other words, the latency of the polling loop signals may prevent the use of closed loop power control. However, the described example allows for a power control sequence that may be effectively carried out in a relatively short span of time, thereby allowing closed loop power control. Preferably, the elapsed time encompassing transmission of the control pulse 215, the base transmission 204, and the start of the user transmission 206 is kept relatively short (e.g., less than 500 µsec or roughly 2.5% of the duration of the major frame 201), allowing system response to be fast enough to counteract small scale multipath fading effects and propagation shadow effects.

The base station 104 may also use the control pulse 215 to measure the time delay from a user station 102 and thereby estimate the distance of the user station 102 from the base station 104. For 911 support, a user station 102 can provide control pulses 215 to multiple base stations 104 for rough location estimation in emergency situations.

The base station 104 may have a plurality of antennas for reception and transmission on the communication link with the user station 102, and may select one of that plurality of antennas for reception and/or transmission, in response to the determination the base station 104 may make in response to the control pulse 215. The base station 104 may make the determination of which antenna to use based on the quality of the signal received from the control pulse 215 transmitted by the user station 102. Because the base station can both receive and transmit on the antenna having the best received signal quality from the control pulse 215, the user stations 102 benefit from antenna selection diversity even though they might not have explicit antenna diversity capabilities at the user station 102. The control pulse 215 permits spatial diversity control to be updated during each minor frame 202. Preferably, the base station 104 employs a high speed TDD technique such that the RF channel characteristics do not change within the time of the minor frame 202.

Information relating to the control pulse 215 for a particular user station 102 may be transferred as information in control traffic from one base station 104 to another base station 104 in the case of a base station assisted handoff.

It should be noted that, in the preferred TDMA system described herein, the requirement of strict RF transmitter output power control is not necessary to resolve the "near-far" problem commonly experienced in CDMA systems. The purpose of the control pulse 215 is primarily to reduce battery consumption in user stations 102, to minimize interference of transmissions among neighboring cells 103 which may be operating on the same or adjacent RF channels, and to minimize interference with nearby fixed microwave users.

The control pulse 215 may also serve as a synchronization preamble for determining the beginning of M-ary data symbols within the minor frame 202. A power control command pulse, similar in length to the control pulse 215, transmitted by the base station 104 during the base transmission 204 or otherwise may likewise be used as a synchronization preamble at the user station 102, in addition to providing a power control command to adjust the power output level at the user station 102.

Base Station Output Power

Because a single base station 104 may communicate with a large number of user stations 102 (e.g., as many as 64 user stations 102) at a given time, each of whose distance from the base station 104 may vary from near zero up to the radius of the cell 103, it may not be practical to control the transmitter power of the base station 104 in order to maintain a near-constant received power level at each user station 102 during each minor frame 202. Output power control of the transmitter at the base station 104 could require a large change (e.g., more than 40 dB) in transmit power during each minor frame 202 (e.g., every 625 µs) of the major frame 201. As an alternative to providing power control on a minor frame 202 by minor frame 202 basis, output power control at the base station 104 can be averaged over a longer time interval than each minor frame 202.

Antenna Characteristics.

The reciprocal nature of time division duplex (TDD) permits common antennas to be used for transmit and receive functions at both the base station 104 and the user stations 102, without the need for antenna diplexers. Common antennas can be used to transmit and receive because these functions are separated in time at each of the terminals. Further, because TDD utilizes the same RF frequency for the transmit and receive functions, the channel characteristics are essentially the same for both the base station 104 and a particular user station 102.

The use of common antennas results in simplicity of the base station 104 and user station 102 terminal designs. Further, use of the same RF frequency and antenna for both transmit and receive functions at the base station 104 and the user station 102 provides reciprocal propagation paths between the base station 104 and user station 102 terminals. This reciprocal nature allows the base station 104 to use the channel sounding of the control pulse 215 transmitted by the user station 102 to determine the two-way path loss between the base station 104 and the user station 102, and also to determine which of the spatial diversity antennas at the base station 104 to use, both to receive from the user station 102 and to transmit to the user station 102.

Different types of antennas may be used by the base station 104, depending on the type of application. For low density suburban or rural applications an omnidirectional antenna may be used to provide maximum coverage with the fewest base stations 104. For example, an omnidirectional antenna may be employed having a vertical gain of approximately 9 dB. The 9 dB of gain permits a relatively large radius cell even with an omnidirectional horizontal pattern.

In suburban and low density urban areas, directional antennas with 120 degree azimuth beamwidths and 9 dB vertical gain may be used at the base station 104 so that a cell 103 can be sectorized into three parts, with each sector accommodating a full load of user stations 102 (e.g., 32 full duplex user stations 102).

The use of TDD also permits utilization of a single steered phased array antenna at the base station 104 for applications requiring a high gain, highly directional antenna. Similar deployment in CDMA or FDMA systems would, in contrast, be more complex and costly, as they may require simultaneous steered beams for each user station 102 within the cell 103.

For example, to permit a single base station 104 to cover large, sparsely populated area, a steered array antenna with up to 20 dB of horizontal directivity can be used. Such an antenna is sequentially steered to each user station 102 within a cell 103 at each minor frame 202. The same antenna may be used for both transmission and reception, as noted, providing reciprocal forward and reverse link propagation characteristics. The steered array antenna may utilize circular polarization so that high level delayed clutter signals reflected from buildings or other obstructions within the beam path do not interfere with the received signals from the user stations 102. As reflected signals are typically reversed in polarization, they will be rejected by the circularly polarized antenna. It should be noted that such high gain, directional antennas also reduce the delay spread in severe multipath environments by rejecting multipath components arriving from outside the main beam of the antenna.

The user station 102 employs a halfwave dipole antenna which is linearly polarized and provides a gain of 2 dB with an omnidirectional pattern perpendicular to the antenna axis. At a nominal frequency of 1900 MHz, a half wavelength is approximately 3 inches, which fits well within a handset envelope.

MESSAGE TYPES AND PROTOCOL

Figure 3 shows message types and a protocol which uses those message types.

Messages (base transmissions 204 and user transmissions 206) may be one of three types: a general poll message 301, a specific poll message 302, and an information message 303. When a message is transmitted by a user station 102, it is called a "response", e.g., a general poll response 304, a specific poll response 305, and an information response 306.

User Station Initiation of a Link

A user station 102 may "acquire" a base station 104 by a sequence of handshaking steps. At a general poll step 307, the base station 104 may transmit its general poll message 301 on an air channel 203 as part of a minor frame 202. The user station 102 receives the general poll message 301 and, if and only if it was received without error, transmits its general poll response 304 on the same air channel 203. The general poll message 301 comprises a base ID 308, which may be 32 bits long, which may be recorded by the user station 102. In like manner, the general poll response 304 comprises a user ID 309, which may be 32 bits long, which may be recorded by the base station 104. The base ID 308 may be used during handoff, as noted herein.

Upon receiving a general poll response 304, at a specific poll step 310, the base station 104 may transmit a specific poll message 302, comprising the user ID 309 received by the base station 104 as part of the general poll response 304. The specific poll message 302 may be transmitted on the same air channel 203 as the general poll message 301, or may be transmitted on another air channel 203, so long as the user station 102 is able to find it.

The user station 102 may monitor all air channels 203 for its specific user ID 309. The user station 102 receives the specific poll message 302 and, if and only if it was received without error and with the same user ID 309, transmits its specific poll response 305 on the same air channel 203. The specific poll response 305 comprises the same user ID 309 as the general poll response 304.

However, the specific poll message 302 may be eliminated as redundant. The user station 102 may therefore follow the general poll response 304 with a specific poll response 305 on a selected air channel 203. This air channel 203 may be designated by the base station 104 in a part of the information field 209 of the general poll message 301, it may be designated by the user station 102 in a part of the information field 209 of the general poll response 304, or it may be selected by the user station 102 in response to an unoccupied air channel 203 (e.g., the user station 102 may seize an unoccupied air channel 203). The latter of these three alternatives is presently preferred by the inventors.

Upon receiving a specific poll response 305 comprising a user ID 309 which matches that of the general poll response 304, at a link-established step 311, the base station 104 may transmit an information message 303. At this point, the base station 104 and user station 102 have established a communication link 312 on a designated air channel 203, typically the air channel 203 originally polled by the base station 104, but possibly a different air channel 203. The base station 104 may couple a telephone line to that air channel 203, and the user station 102 may begin normal operation on a telephone network (e.g., the user station 102 may receive a dial tone, dial a number, make a telephone connection, and perform other telephone operations). The base station 104 and user station 102 may exchange information messages 303 and information responses 306, until the communication link 312 is voluntarily terminated, until faulty communication prompts the user station 102 to re-acquire the base station 104, or until handoff of the user station 102 to another base station 104.

Should more than one user station 102 respond to a general poll message 301 in the same minor frame 202, the base station 104 may advertently fail to respond. The lack of response from the base station 104 signals the involved user stations 102 to back off for a calculated time interval before attempting to acquire the same base station 104 using the general poll message 301 and general poll response 304 protocol. The back-off time may be based upon the user ID 309, and therefore each user station 102 will back off for a different length of time to prevent future collisions.

The general poll message is sent by a base station 104 on one or more currently unoccupied air channels 203. Originally, at power-up of the base station 104, the base transmission 204 for all of the air channels 203 may therefore contain the general poll message 301.

Base Station Initiation of a Link

When an incoming telephone call is received at a base station 104, at an incoming-call step 313, the base station 104 transmits a specific poll message 302 with the user ID 309 of the indicated recipient user station 102 (skipping the general poll message 301 and the general poll response 304) on an available air channel 203.

Each user station 102 listens for the specific poll message 302 repeatedly on each air channel 203 so as to receive the specific poll message 302 within a predetermined time after it is transmitted. Thus each user station 102 may periodically receive each air channel 203 in sequence so as to listen for the specific poll message 302.

When the specific poll message 302 is received, the user station 102 compares the user ID 309 in the message with its own user ID, and if they match, continues with the link-established step 311. The base station 104 may thus establish a communication link 312 with any user station 102 within communication range.

Link Expansion and Reduction

The data transmission rate between a base station 104 and a user station 102 may be expanded or contracted over the duration of the communication link.

The base station 104 increases the data transmission rate by transmitting multiple information messages 303 to the user station 102 during a major frame 201, essentially allocating multiple minor frames 202 to a single user station 102. These higher data rates, also known as "super rates", are implemented by means of a targeted information message 303. In a targeted information message 303, the base station 104 may transmit the user nickname 212 in the D field 208, along with information to be transmitted to the designated user station 102 in the B field 209. When the user station 102 detects the user nickname 212 assigned to it, it receives the targeted information message 303.

The user nickname 212 may be transmitted by the base station 104 to the user station 102 in the specific poll message 302. In an embodiment where the specific poll message 302 has been eliminated as redundant, the user nickname 212 may be transmitted by the base station 104 to the user station 102 bit-serially in a designated bit of the header field 207.

Because the data transmission rate is related to the number of minor frames 202 allocated to a specific user station 102, the data transmission rate increases in steps of, for example, 8 Kbps. It is contemplated that up to the full bandwidth of the base station 104 -- that is, up to all 32 full duplex slots or 256 Kbps (full duplex) - - may be assigned to a single user station 102.

Data rates lower than the basic rate (i.e., less than one minor frame 202 per major frame 201 or less than 8 Kbps) are provided. The lower data rate is accomplished by skipping major frames 201 on a periodic basis. Thus, data rates such as 4 Kbps, 2 Kbps, and so on can be provided. In one embodiment, up to 24 consecutive major frames 201 may be skipped, providing a minimum data rate of 320 bps efficiently (i.e., without using rate adaptation). Intermediate rates or even lower rates may be obtained by using rate adaptation.

The capability of providing variable data rates on demand, including availability of an asymmetric mode in a given minor frame 202 described earlier, provides an efficient and flexible data conduit for a wide array of data, video, multi-media and broadcast applications. For example, each minor frame 202 can be configured with the majority of the minor frame 202 duration allocated to either the base transmission 204 or the user transmission 206, or can be configured with a symmetric distribution in which half of the minor frame 202 duration is allocated to both the base transmission 204 and the user transmission 206. Typically, voice traffic utilizes a symmetric distribution as either end of the link may send voice traffic. In a data exchange, however, more data is typically sent in one direction and less in the other. For instance, if fax data is being sent to a user station 102, then a higher data rate for the base transmission 204 would be advantageous and is supportable with the described configuration. For even higher data rate applications, a particular base station 104 or user station 102 may be assigned multiple minor frames 202 within a single major frame 201. These high data rate modes can support, for example, enhanced voice quality, video data or broadcast data applications.

Handoff and Network Maintenance

Once a base station 104 and user station 102 have established a communication link 312, during the link-established step 311 the user station 102 may receive all information messages 303 and transmit all information responses 306 on the same air channel 203 or on specified multiple air channels 203. This arrangement leaves the remainder of the major frame 201 free for other activities. One such activity is to interrogate other base stations 104 and maintain network information such as link quality and channel availability at nearby base stations 104 in order to facilitate handoffs from one base station 104 to another base station 104.

Base stations 104 transmit network information as part of the general poll message 301 and the specific poll message 302, in a channel utilization field 314 or otherwise. The network information may include, for example, the identity of nearby base stations, the identity or relative amount of free channels at a particular nearby base stations and/or at the current base station, link quality for nearby base stations and/or the current base station, and frequencies and spread spectrum code sets used by the nearby base stations.

At a network-maintenance step 315, the user station 102 may listen on one or more different air channels 203, other than the one(s) currently being used by the user station 102, for the general poll message 301 and the specific poll message 302 from nearby base stations 104. The user station 102 continues to communicate on its designated air channel(s) 203 with its current base station 104 and responds as necessary to information messages 303 from that base station 104. However, unless a handoff procedure is initiated as described below, the user station 102 does not transmit in response to other nearby base stations 104 and therefore does not occupy air channels 203 of those base stations 104.

It is contemplated that the system may perform either a "make before break" handoff for seamless, undetectable handoffs, or a "break before make" handoff in emergency situations where all communications with a base station 104 are lost prior to a new connection being established.

In a "make before break" handoff, if the communication link 312 between the base station 104 and the user station 102 is too faulty, then the user station 102 may acquire one of the nearby base stations 104 in like manner as it acquired its current base station 104. Such a handoff procedure may be further explained with reference to Fig. 1-3.

In Fig. 1-3, it is assumed that a user station 102 presently in communication with a current or original base station 405 has determined it to be desirable to transfer communication to a different base station 104, such as a first terminal base station 410 coupled to a common base station controller 407, or a second terminal base station 406 coupled to a different base station controller 408. A handoff to the first terminal base station 410 will be termed an "intra-cluster" handoff, while a handoff to the second terminal base station 406 will be termed an "inter-cluster" handoff. The following explanation will focus on an intra-cluster handoff to the first terminal base station 410, but many of the steps are the same as with an inter-cluster handoff, and the salient differences between an intra-cluster and inter-cluster handoff will be noted as necessary.

In general, when the user station 102 determines that a handoff is appropriate, the user station 102 acquires an air channel on the new or terminal base station 410 and notifies the base station controller 407 coupled to the current base station 405 to switch the incoming phone line from the current base station 405 to the new base station 410.

More specifically, a handoff procedure may be initiated when the received signal level at a user station 102 falls below an acceptable level. While the user station 102 receives bearer traffic from its originating base station 405, the user station 102 measures the received signal quality (e.g., RSSI) of its communication link 312. The received signal quality value, together with measurements of the current frame error rate and type of errors, determines the overall link quality. If the overall link quality drops below a first threshold (the measurement threshold), the user station 102 begins searching for available air channels 203 (i.e., time slots), first from the originating base station 104, and then (using appropriate frequencies and spread spectrum codes) from neighboring base stations 104 of adjacent or nearby cells 103. The user station 102, as mentioned, preferably has obtained information regarding the identities of neighboring base stations 104 (including spread spectrum code set and frequency information) from the originating base station 405 by downloading the information to the user station 102 during traffic mode or otherwise.

As the user station 102 scans potential new air channels 203 using the appropriate frequency and/or spread spectrum code set, the user station 102 measures and records the received signal quality. The user station 102 reads a field carried in all base transmissions 204 which describes the current time slot utilization of the base station 104. The user station 102 uses these two pieces of information to form a figure of merit for the new base station signals, including the originating base station 405, and then sorts the base stations 104 by figure of merit. This procedure allows the user station 102 to evaluate the quality of available air channels 203 for both the originating base station 405 and other nearby base stations 104.

If an air channel 203 (or air channels 203, as the case may be) for the originating base station 405 has better quality than that of any base station 104 in adjacent or nearby cells 103, a time slot interchange (TSI) handoff is considered, which maintains the link to the originating base station 405 on a different air channel 203 than was previously being used by the user station 102.

If the link quality drops below a second threshold level, then the user station 102 (during a no-bearer time slot) requests a handoff from the base station 104 with the highest figure of merit (which could be a TSI handoff with the originating base station 405). The handoff is requested by seizing an air channel 203, sending a handoff message request, and waiting for an acknowledgment from the new base station 410. The handoff signaling message contains a description of the circuit connecting the originating base station 405 to the network, which description was passed to the user station 102 at call establishment time. If the new base station 104 accepts the handoff request (by acknowledging), then the new base station 104 becomes the terminal base station 410. Note that the user station 102 maintains its original air channel 203 connection with the originating base station 405 during this handoff procedure, at least until a new air channel 203 is acquired.

To complete an intra-cluster handoff, at a handoff step 316 the user station 102 transmits to the new base station 410 the base ID 308 of the old base station 405. The old base station 405 and new base station 410 may then transfer the handling of any telephone call in progress.

More specifically, the terminal base station 410 sends a message in the form of a "note" (as previously described) to its base station controller 407, requesting that the original circuit be switched from the originating base station 405 to the terminal base station 410. If the base station controller 407 is common to both the originating base station 405 and terminal base station 410, the handoff is termed an intra-cluster event, and the base station controller 407 bridges the circuit from the originating base station 405 to the terminal base station 410. The base station controller 407 then sends a circuit-switch-complete note to the originating base station 405 and also to the terminating base station 410, commanding the latter to continue the handoff process.

In the case of an inter-cluster handoff, the base station controller 408 is not common to both the originating base stations 104 and the terminal base station 406. For these types of handoffs, as with intra-cluster handoffs, the terminal base station 406 sends a message in the form of a note to its base station controller 408, requesting that the original circuit be switched from the originating base station 405 to the terminal base station 406. The base station controller 408 translates the handoff note into the signaling language of the network host 409 (e.g, a PCSC) and requests an inter-cluster handoff at the network level.

In some network architectures, the host network 409 cannot accept a handoff request from a terminating base station controller 408, in which case an intermediate step is taken. The handoff request may be sent via an X.25 link to the base station controller 407 connected to the originating base station 405. The originating base station controller 407 then translates the handoff request and relays it to the network host 409. The network host 409 acknowledges the circuit switch to the originating base station controller 407, which then sends a circuit-switch-complete note to the terminal base station 406.

When the terminal base station 406 receives the circuit-switch-complete note, the terminal base station 406 begins paging the user station 102 with a specific poll, and the originating base station 405 signals the user station 102 to transfer to the terminal base station 406. When the user station 102 receives the signal to transfer to the terminal base station 406, or if the link is lost during the handoff process, the user station 102 switches to the terminal base station 406 and searches for a specific poll message 302. When the user station 102 receives the specific poll message 302, the user station 102 completes the connection to the terminal base station 406, and the handoff procedure is finished.

Should the link between the user station 102 and the originating base station 405 or terminating base station 406 (or 410) be completely broken at any time, the user station 102 will search for the highest quality base station 104 on its list of potential handoffs, and attempt a handoff without communication with its previous base station 405. This capability allows the user station 102 to recover from situations in which the original link was broken before the normal handoff procedure could be completed.

An intra-cluster handoff, including re-establishment of bearer channel traffic, may ordinarily take from less than 10 milliseconds to as much as 40 milliseconds. Since under normal circumstances the handoff time is less than one polling loop interval, bearer packets will continue to the user station 102 with no interruption. Inter-cluster handoff times are partially dependent upon the delays inherent in the host network 409 and are not always easily predictable.

A unique aspect of the above described "mobile directed" or "mobile centric" handoff technique is that the user station 102 makes the decision to handoff between cells and directs the base station controller or network to make a line switch once an alternative base station 104 is acquired. This approach is quite different from a "network directed" or "network centric" approach such as used in systems such as AMPS, IS-54 cellular, and GSM. The mobile centric approach also differs significantly from so-called "Mobile Assisted Handoff" (MAHO) in which the network collects information and directs all or most of the handoff functions, thereby utilizing the user station 102 primarily as an additional listening post with the network still directing the handoff. The MAHO technique therefore ordinarily requires significant signaling and messaging between base stations, base station controllers, and switches, causing handoffs to take much longer than with the mobile centric techniques described herein.

A major benefit of the mobile centric approach is that it may allow for mobile speed handoffs (e.g., 65 MPH) even in very small or very large cells, such as cells ranging from as small as under 1000 feet to as large as 20 miles in diameter.

The system is also capable of performing a "break before make" type of handoff as well. A "break before make" handoff is typified in a situation where sudden shadowing occurs, such as when a connection with the current base station 405 is lost due to a severe signal blockage (e.g. worse than 40 dB) near the limit of the cell range such as can occur when turning a corner quickly in a dense urban high rise area. In such a situation, the user station 102 checks its previously created "priority list" of available base stations in the vicinity and attempts to establish contact with a new base station 104, perhaps on a new frequency and/or a new time slot. The user station 102 may include as part of its control logic a "persistence" parameter which will preclude call tear down from occurring before a duplex connection is fully reestablished.

The true "hard handoff" problem (i.e., a lost air channel) may in many instances be handled very quickly through the ability of the user station 102 to re-acquire the original base station 405 or to acquire a different base station 104 very rapidly even when no information is available to the user station 102 when the link was lost. Even in such an emergency "break before make" handoff situation, the handoff may ordinarily be accomplished in as little as 16 to 250 milliseconds. In contrast, complete loss of a link in traditional cellular architectures becomes a "dropped call."

One problem that may occur during handoff is a situation in which there are repeated attempts to switch between two or more base stations 104 during times, for example, when the measured quality of the received signals from two competing base stations 104 is very close, or when environmental effects cause rapidly changing deviations in the relative measured signal quality of the signals from competing base stations 104. The repeated switching between competing base stations 104 may be referred to as "thrashing" and may have the undesirable effect of consuming excess capacity from the network. In order to reduce the effect of thrashing, hysteresis measurements from multiple base stations 104 may be maintained by the user station 102 so that a handoff does not occur until the quality of the signal from a new base station 104 exceeds the quality of the signal of the original base station 405 by a predetermined margin. In such a manner, important air channel resources in the network may be preserved.

In rare instances, two user stations 102 on the same minor frame 202 in different cells 103 but on the same frequency may encounter propagation characteristics in which the spatial and code separation are insufficient to prevent bit errors, thus causing the user stations 102 to begin experiencing degradation of their RF links. In such cases, a time slot interchange (TSI) may be performed wherein one or both of the conflicting user stations 102 are assigned different minor frames 202 within their respective major frames 201 to eliminate further collisions. Such a procedure may be viewed as the time domain equivalent of dynamic channel allocation as the system either assigns an unoccupied air channel 203 to the user station 102 or switches the user station's 102 minor frame 202 with that of another user station 102 in the same cell 103 which is geographically removed from the interference.

Security and Error Handling

The protocol protects communications against errors in several ways: protocol handshaking, user ID verification and reverification, and synchronization by reacquiring the base station. Handshaking, verification and synchronization protect both the base station 104 and the user station 102 from receiving telephone calls in progress on any other air channels 203.

Handshaking provided by the general poll step 307 and the specific poll step 310 requires that the proper message having the proper header be transmitted and received, and in the proper sequence. In each message, the header field 207 (sixteen bits) is protected by a CRC code 211 (four bits); an error in the header field 207 or in the CRC code 211 indicates an error and will cause the protocol to restart handshaking with the general poll step 307.

The user ID is verified twice, once by the base station 104 and once by the user station 102. In the general poll message 301 and specific poll message 302, the user ID 309 is protected by a CRC code 211 (sixteen bits), in like manner as the CRC code 211 for the header field 207. An error in the user ID 309 or in the CRC code 211 will cause the protocol to restart handshaking with the general poll step 307.

At the link-established step 311, the base station 104 and the user station 102 are protected against drift and/or desynchronization, even when transmission or reception are interrupted. When a threshold for an error rate is exceeded, the base station 104 and user station 102 each independently stop sending data in information messages 303 and information responses 306, and return to the specific poll step 310 for resynchronization. When the specific poll message has been eliminated as redundant, the base station 104 and the user station 102 may determine resynchronization by means of a designated bit in the header field 207.

At the specific poll step 310, the base station 104 transmits the specific poll message 302 and the user station 102 searches the major frame 201 for a specific poll message 302 having a user ID 309 which matches its own user ID 309. After this handshaking succeeds, the base station 104 and user station 102 return to the link-established step 311 and continue transmitting and receiving information messages 303 and information responses 306.

This technique for recovery from desynchronization, also called "reacquiring the base station," has the advantage that both the base station 104 and the user station 102 independently reverify the user ID 309 before communication is resumed. This assures that the base station 104 and the user station 102 stay in synchrony and communicate only on the agreed air channel 203. Should the base station 104 and the user station 102 be unable to reestablish the communication link 312, the telephone call will be terminated by the base station 104.

At the link-established step 311, the base station 104 also repeatedly and periodically transmits the user ID 309 in the D field 208 of the information message 303. The user station 102 checks the user ID 309 to assure that the base station 104 and the user station 102 are each communicating on the proper air channel 203. If this user ID 309 does not match, it returns to the specific poll step 310 to reacquire the base station 104, as noted above.

Protocol Flexibility

The protocol described above provides flexibility with a small number of unique messages. The protocol is immune to changes in polling loop length and in the number of air channels allowed. The number of simultaneous users is therefore responsive to voice compression and data rate constraints and not by the protocol. The protocol also provides for an unlimited number of user stations in a given area, with the provision that the number of simultaneous calls cannot exceed the number of air channels. An unlimited number of base stations are also supported, making base station geography a function of available frequencies and range, not of protocol. The ability to interrogate and acquire alternate base stations in the presence of faulty communication provides for the expansion of a microcell network which may use base station handoff to route calls to base stations within range.

System Synchronization

In order to maximize system throughput capacity, the TDMA frame times for all base stations 104 within a geographical region are preferably synchronized to within a specified tolerance. For example, all base stations 104 begin transmissions for the same frame within 6 microseconds.

The primary data timing standard in a digital network backhaul system, such as T1, ISDN BRI, or PRI, is the public switched telephone network (PSTN) timing standard. To prevent data precession into over run or under run, all base station controllers 105 and base stations 104 in such systems are synchronized to the PSTN timing standard.

At the system level, a GPS receiver is used at each base station controller 105 (and optionally at each base station 104) to generate the primary reference timing marker for the TDMA frame timing. This marker is captured at the base station controller 105 every second and transmitted to the attached base stations 104. A base station controller may temporarily turn off any major frame 201 or minor frame 202 of a given cell 103 which may be interfering with a neighboring cell 103.

Each base station 104 provides the basic TDMA loop timing structure for its cell or sector. As previously noted, a synchronization preamble in the form a control pulse 215 or power control command is transmitted at the beginning of each minor frame 202 by the user station 102 and the base station 104, respectively. When the appropriate preamble, consisting of a code sequence 48 chips in length, is received, a digital correlator (i.e., a matched filter) attuned to the specific preamble generates an internal synchronization pulse which may be very brief (e.g., two chips in duration, or 400 nanoseconds). The internal synchronization pulse may then be used to synchronize the start of M-ary symbol detection process.

The system referred to as the Omnipoint system is composed of the following four network hierarchical elements; Mobile Station (MS), Base Station (BS), Base Station Controller (BSC), and a Personal Communications Switching Center (PCSC).

Mobile Station (MS)

The Mobile Station (MS) is the subscriber terminal. It communicates with one Base Station (more during handover) and is responsible for the following:

Mobility Management. Unlike most full mobility systems, the Omnipoint MS controls the handover actions of the system. It is responsible for monitoring the quality of the ongoing bearer connection with its current Base Station as well as the quality of signals received from other nearby Base Stations/Cells. If the quality of the bearer channel drops below an acceptable level the MS will initiate a handover to a better BS. The MS is also responsible for maintaining registration with its current BS as well as directing the BSC to make a switch of BS's as part of the handover.
Call Control. The MS is responsible for initiating outgoing call requests and periodically monitoring for incoming call pages. Page monitoring is scheduled on duty cycle basis to reduce power consumption.
Transmission. The MS maintains the over-the-air channel from the mobile perspective. It performs all necessary timing, slot aggregation and error recovery.
Authentication. The MS maintains the subscriber secret key and performs the required calculations necessary to authenticate the device.
Applications. The MS executes the appropriate downloadable user applications. The applications implement the short messaging service, broadcast messaging and data services. Additional user specific applications using the system data services may also be downloaded and executed.
Subscriber Interface. The MS implements the feature activating subscriber interface as well as the interface to specific downloaded applications in the MS.

Base Station (BS)

The Base Station (BS) is an intelligent device which controls the over the air signal and serves as Cell controller. It communicates with multiple MSs and a Base Station Controller. It provides the following functionality:

Radio Resource Management. The BS controls the over-the-air slots of the TDMA spread spectrum transceiver. The BS allocates these slots based upon MS initial service requests, handover bandwidth demands, special service needs and OA&M test and operational restrictions. In addition, the BS is the component which is charged with allocation of specific slots, super-slots, and sub-slots to maintain the requested bandwidth.
Mobility Management. The BS responds to MS handover requests if resources are available. The BS assigns network and radio resources to the MS and sends a request to the Base Station Controller to execute a handover switch. The BS also maintains registrations with MSs within its Cell. This registration information is maintained in a stub or cache Visitor Location Register (VLR).
Time Slot Interchange (TSI). This function allows a Mobile or a Base Station to change the operational time slot if interference is detected, taking advantage of statistical placement of Mobile Stations in both the original and re-used cells. The system supports very quick TSI, within a single polling loop (i.e., less than 20 msec), as required and available.
Call Control. The BS interprets the over-the-air signaling traffic and translates this traffic from its concise form to an appropriate network independent format called Notes. These Notes have been designed to reduce network latency and bandwidth while communicating succinct information to the higher elements in the network.
Transmission. The BS is responsible for the allocation of network channels to their corresponding air slots. The BS controls these resources for both incoming and outgoing calls. Audio connections over network channels are maintained in their compressed form. In order to support a wide variety of bandwidths to the MSs, multiple network time slots may be assigned to a specific MS. The BS is also responsible for monitoring the interference on the radio and adjusting its slot utilization to minimize this interference. In addition, the BS measures signal quality from its MSs and selects antennas for space diversity and polarization when appropriate, as well as controls the MS power.
Authentication. The Base Station responds to Notes from the Base Station Controller to authenticate a given MS. The BS sends the appropriate over-the-air signals to the MS and compares the authentication response against the proper value.
Operations, Administration and Maintenance. The Base station maintains accounting information for each connection in an accounting log. This information along with maintenance statistics are transferred to the Base Station Controller and higher level entities via Notes. Notes are also used to block, unblock, and loopback-test network channels for maintenance purposes. The PCSC also provides OA&M.

Base Station Controller (BSC)

The Base Station Controller (BSC) is an intelligent switch which controls and performs switching functions for a group of Base Stations called a cluster. It communicates with multiple Base Stations, and a PCSC. The Base Station Controller functionality is:

Mobility Management. The BSC interprets handover requests from its Base Stations. If the handover request represents an intra cluster handover, i.e. between two Base Stations on the same BSC, the BSC switches all of the circuits associated with the MS from the originating BS to the terminating BS. If the handover request indicates that the originating BS is located on some other BSC then the handover Note is translated into a form appropriate for the host PCSC and this form is forwarded to the host PCSC.
Call Control. The BSC is responsible for cluster location of the MSs. Registration information forwarded from a BS is recorded in a local stub Visitor Location Register (VLR) located in the
BSC. This is used to locate the MS during alerting operations. Line resource requirements forwarded from the BS are also held in this stub VLR. Furthermore, the BSC translates the commands and data contained in Notes passed to it by the BS into signaling suitable for the host PCSC system. In this sense the BSC acts as a translator between the Omnipoint system and the host system.
Transmission. The BSC is responsible for allocation and termination of the trunking circuits between the BSC and the host PCSC. The BSC is also responsible for the transcoding of compressed audio to the trunking format. The BSC further serves as an Inter-Working-Function between the Omnipoint data channelization format and other network formats.
Operation, Administration and Maintenance. The BSC stores accounting information delivered to it by the BSs. This information is translated into a form suitable for the host system and forwarded to the PCSC. In addition the BSC is responsible for blocking, unblocking and loop testing circuits upon request from the PCSC.

Personal Communications Switching Center (PCSC)

The PCSC is the primary switch and the interconnect to the PSTN. The Omnipoint system has been designed to connect to a variety of existing and proposed switch systems, including the AIN network, the Americanized version of GSM, IS-41, and ATM-based networks. The operation of these systems is outside the scope of this response.

Mobility Management. The PCSC provides inter cluster and inter PCSC handover capability. In addition the PCSC provides Home Location Register (HLR) and Visitor Location Register (VLR) services. In this capacity the PCSC responds to its normal signaling messaging as provided by the BSC translation function.
Call Control. The fundamental operations necessary for call control, routing, and call features are supported by the PCSC.
Transmission. All PSTN transmission facilities are managed by the PCSC.
Authentication. Authentication requests and data are initiated by or stored in the PCSC.
Operations, Administration and Maintenance. All OA&M functions are initiated by the PCSC.

The Omnipoint system utilizes a unique combination of Frequency, Time, and Code Division Multiplexing. Within a cell, Time Division Duplexing (TDD) and Time Division Multiple Access (TDMA) are employed, allowing up to 32 simultaneous, 8 Kbps full duplex users or 64 full duplex 4 Kbps users, while adjacent cells are set to different frequency channels (FDMA) under a minimal N=3 architecture. Cells beyond adjacent cells use a variety of multiplexing tools and capabilities including code (CDMA) and time slot (TDMA) separation and Time Slot Interchange (TSI) for further inter-cell multiplexing isolation,. Under certain circumstances, Voice Activity Detection (VAD) can be used to increase the number of voice users (theoretically, up to twice the half-duplex number). Going to 4 Kbps also allows up to twice the number of voice users without any changes to the CAI or the Base Station hardware.

The TDD/TDMA structure is based on a 20 ms polling loop which supports thirty-two, 8 Kbps full duplex time slots of 625 µsec duration and provides that each Mobile Station be able to aggregate multiple slots to afford a user more or less data bandwidth as required. Asymmetric data rates can be supported on a frame by frame basis, thus enabling VAD, high speed data transport, and data broadcasting. Thus, for broadcast applications, for instance, all the data can be set up to emanate from the base to the Mobile Station allowing 16 Kbps data transfer rate in one slot, and with slot aggregation up to 512 Kbps.

All full duplex transmissions from BS to MS, and from MS to BS are synchronous (See Section 4.17), such that from cell to cell, Base Stations transmit during the allotted portion of the slot, and Mobile Stations transmit during their allotted time within the slot (See Figure 4). Within each cell, only the BS or one of many Mobile Stations transmits at any instant in time.

Intra-Cell Multiplexing. Since each Mobile Station within a cell is Time Division Duplexed, and Time Division Multiplexed with all other Mobile Stations in the same cell through a synchronous slot structure controlled by the Base Station, they are not hindered by positional (near/far) restrictions within the cell, providing "perfect" time isolation between mobiles.
Inter-Cell Isolation & Re-Use. On an Inter cell basis, a series of multiplexing tools and capabilities, including TSI, power control, code orthogonality, antenna diversity and beam directivity, supply significant isolation from cell to cell.

Under most "real world" scenarios, propagation characteristics and significant shadowing (especially at 1.8 GHz) alone will most often provide sufficient cell isolation. Additionally, the system takes advantage of mechanisms that provide significantly higher degrees of isolation between re-used cells, including:

Time Slot Interchange (TSI). This function allows a Mobile or a Base Stations to change the operational time slot if interference is detected, taking advantage of statistical placement of Mobile Stations in both the original and re-used cells. The system supports very quick TSI, within a single polling loop (i.e., less than 20 msec), as required and available.
Transmitter Power Control. By controlling the Mobile Transmit Power in discrete steps (minimum 3 dB each), the minimal emitted RF power reduces interference into all cells. The Time Slot structure allows for power adjustment in less than 500 µsec, which will mitigate both interference and fading events.
Antenna Diversity and Directivity. The system supports both antenna diversity and directional antennas, whether they are steered phased array or sectorized architectures. With the system supporting phased array antennas, highly directional RF beams may be utilized that dramatically decrease interference to other cells and reduce delay spread.

Time Slot Interchange, Power Control, Antenna Diversity and Directivity capabilities, coupled with shadowing, propagation, and statistical location effects, will provide dramatic isolation gains between Base Station and Mobile Stations sharing the same frequency channel under the reuse plans. Additionally, the Omnipoint system obtains further gains when utilized in an environment that provides more than three discrete frequency channels.

TDMA Access Structure Polling Loop Description

The Omnipoint polling loop is composed of 32 full duplex 8 Kbps time slots. Each slot is formed of one of three frame types;

A) a symmetric base to MS frame and a MS to base frame,
B) an asymmetric base to MS frame and the MS to base frame with acknowledgment, or
C) a single simplex broadcast frame without acknowledgment. Non-broadcast frames can be in either poll or traffic mode.

Additionally, in the Omnipoint system, an ISDN-like "D" channel format is provided in every packet for application-specific information. This information can be used for signaling, short message service (such as in GSM), voice mail notifications, paging during communications, or other data messaging applications. At the present time, the D channel is available in its entirety to the user and is not used for any call processing or Omnipoint system information. LAPD or Q.921-like error correction algorithms are used with data transmission via the D channel to ensure delivery and acknowledgment of the information.

Packet Contents

The following is an illustration of an Omnipoint packet that could either contain either normal or control traffic.

The header in the packet identifies the Base Station or MS, packet type, link quality, and other information required by the system for efficient operation. The packet also contains signaling and/or messaging information in the D channel and the bearer information.

Multiplexing and Cell Separation

The Omnipoint system uses a combination of CDMA, TDMA, and FDMA for separation of users and cells. Within a cell, a high speed direct sequence spread spectrum TDMA protocol is used for separating users.

SEE FIGURE 2-1

As shown in figure 2-1, adjacent cells are separated by frequency (FDMA) with CDMA being used to improve the re-use pattern to an N=3 in the standard configuration when frequencies are repeated. Additionally, Time Slot Interchange (TSI) and directional antennas are employed to create further isolation of users across cells.

Note that each user within a cell can use the same set of direct sequence codesets. This is contrary to CDMA-only systems that use codes for user separation within a cell. Omnipoint's system uses a different codeset each time the same frequency is reused (noted as C1, C2, and C3 in the preceding picture). The spreading codes are also used within a cell to achieve significantly higher data rates and thus more TDMA users per RF channel than generally thought possible with traditional TDMA systems for fully mobile applications. This direct sequence spread spectrum approach is not a traditional use of CDMA, but is a unique way of using code technology for a wireless system. In the Omnipoint system, code separation is used in combination with spatial and temporal separation to ensure that users on the same frequency in different cells do not interfere with each other.

The most significant benefit of Omnipoint's system architecture is that the cost per user channel declines as the number of users increases per base station, unlike any other system. Generally, in traditional cellular approaches the cost per voice channel remains linear with each additional, simultaneous user. As a consequence, Omnipoint's system offers a five to twenty fold reduction in the cost of providing the voice channels throughout the cell sites.

The Omnipoint approach is also very different from CDMA-only systems that rely on precision power control algorithms and require soft hand-off to achieve high capacity claims and to mitigate near/far problems, resulting in the associated cost and complexity in handsets, base stations and base station controllers. Omnipoint's system could use soft hand-off if cost were unimportant, but does not require it to meet PCS performance and capacity demands. Additionally, with the Omnipoint system the interference generated by an entire cell is never more than a single user, since no more than one transmitter in a cell is on at any instant in time. This is particularly important for frequency planning and OFS coordination in early PCS deployments.

Mobile Access/Link Setup Polling Loop Description

A polling loop, is comprised of 32 Time Division Duplexed (TDD) slots as shown in Figure 2-3. Each time slot is divided into two parts, a BS to MS transmit interval and a MS to BS interval and is structured as sixty four Time Division Multiple Access (TDMA) time slots. Duplexing is accomplished by Time Division Duplexing each TDMA time slot pair. The Slots are not numbered and there is no index associated with the polling loop. Slot synchronization is achieved solely by timing. The first part of a TDD slot is the Base Station (BS) to Mobile Station (MS) transmit frame and the second part of a TDD slot is the MS to BS transmit frame. The BS initiates all transaction between the BS and MS. A General Poll frame is how a MS acquires a connection to a BS. Any time slot that is not seized By a MS contains a General Poll command in the BS to MS part. To acquire a slot, the MS responds to the General Poll Response with a General Poll Response. The BS, upon receiving the General Poll Response, sends a Specific Poll Response. See Figure 2-3. Following signaling data exchanges, the BS and MS begin Bearer Traffic Exchange.

Polling Loop Design - Omnipoint Version 2T.X

A 20 msec loop time is utilized in the Omnipoint Version 2T.X system for a total of 64 8 Kbps time slots simplex, or in a standard voice mode 32 8 Kbps full duplex users per base station RF channel. Figure 2-3 illustrates the loop design.

The Omnipoint system supports two types of polls that can be used for communications between the base station and the handsets within a cell. These comprise the following:

General polls - general polls are used by the base station to allow any handset to communicate with a base station for the purpose of seizing an available air slot or to provide information about the base prior to beginning the handoff process.
Specific polls - specific polls are used by the base station to solicit a response from a specific handset. This type of poll is used by the base station for finding and/or starting communication with a specific handset.

Polls are used by the base station for establishing and maintaining communications with all of the handsets that are within its cell coverage area.

General Poll

The General Poll is a command sent from the BS on any currently unoccupied slot. At BS power-up, all 32 BS transmit slots contain a General Poll. The General Poll is an invitation for any MS to seize the time slot. From a BS software point of view the information in the transmit frame header that defines a General Poll is a bit that signifies that this is a poll type frame, a bit that signifies that the BS is the originator of this frame, command bits that signify that this is a General Poll, and a base ID in the data portion of the frame. When a MS wishes to seize a timeslot, it receives a General Poll in an available time slot and responds, in the MS to BS part of the time slot, with a General Poll Response. From the BS software point of view the information in the General Poll Response header is a bit signifying that this is a poll type frame, no transmission error bit set, and the MS ID (PID) in the data portion of the frame. Upon receiving the General Poll Response, the BS transmits a Specific Poll to the MS. Should more than one MS respond to a General Poll in a particular time slot, the collision and subsequent no response from the BS will cause the involved MS's to back-off for a calculated time interval before attempting to seize a time slot via a General Poll and General Response. The back-off algorithm uses the MS ID and therefore each MS will back-off for a different length of time.

Specific Poll

The Specific Poll is a command sent from the BS to the MS in response to a General Poll Response from a particular MS. See Figure 2-3. The General Poll Response contains the MS's PID. From a BS software point of view the information in the transmit frame header that defines a Specific Poll is a bit that signifies that the BS is the originator of this frame, command bits that signify that this is a Specific Poll, and the MS PID in the data portion of the frame. This PID is issued as a part of the Specific Poll command so that only the originating MS will receive the Specific Poll. Upon receiving a Specific Poll, a MS will respond with a Specific Poll Response in the MS to BS transmit interval of the time slot. From the BS software point of view the information in the Specific Poll Response header is a bit signifying that this is a poll type frame, no transmission error bit set, and the MS ID (PID) in the data portion of the frame matching the PID sent out in the Specific Poll. When the BS receives the Specific Poll Response, it goes through a signaling traffic exchange with the MS, leading directly to Bearer Traffic Exchange.

Channel acquisition occurs whenever a handset needs to access a base slot or requires resynchronization with a base slot. During channel acquisition the handset begins to search for an available base station in its general vicinity by listening for a common signaling poll on any time slot from any possible frequency. Once a base is located with sufficient signal strength and adequate load available, the handset instantaneously seizes an air slot in response to the base station's general poll and the handset provides the base with ID and user information. The base station then transmits a specific poll for that handset to commence communications. At this point, that time slot is removed from being in common signaling mode and the handset and base station are ready to go into traffic mode and begin communications on that time slot. Collisions, during time slot acquisitions, are resolved by an 802.3-like backoff procedure.

The traffic mode in the Omnipoint system occurs after a base station and handset have established a transmission path with the appropriate allocation of air slot resources for the service being supported.

In the traffic mode, two basic types of messages can be supported. These include:

Normal or bearer traffic - bearer traffic consists of the actual information that is being carried by the wireless link, either a voice link or some other form of digital information. Omnipoint supports a bearer channel of 128 bits per frame in each direction (256 bits per frame full duplex).
Control traffic - control traffic provides for link specific data messaging or call control information.

In extremely rare instances, two users on the same time slot in different cells but on the same frequency may encounter propagation characteristics in which the spatial and code separation are insufficient to prevent bit errors and thus the users begin to experience degradation of their radio links. In these cases, the Omnipoint system performs a Time Slot Interchange (TSI) wherein one or both of the conflicting handsets are assigned different time slots within its base station to eliminate the collisions between the two users. This is the time domain equivalent of dynamic channel allocation and the system either assigns a vacant time slot to the handset or switches the handset's time slot with that of another user in the same cell which is geographically removed from the interference. The effect is the removal of interference to the users originally experiencing interference due to poor location specific C/I ratios.

Bearer Traffic Exchange

Once the MS has seized a time slot and been validated (D-channel data), the exchange between the MS and BS leads to Bearer Traffic Exchange (B-channel data). Bearer Traffic can be voice, error-controlled data, or non-error controlled (raw) data. From the BS software point of view the information in the Bearer Traffic header is the absence of a bit signifying that this is poll type frame, no transmission error bit set, and no bit signifying signaling data. Each TDD slot can be configured with the majority of the slot time given to the BS to MS direction, the majority of the slot time given to the MS to BS direction, or symmetric distribution giving fifty percent of the time to both BS and the MS. Typically voice traffic utilizes the bearer channel symmetrically as either end of the link may send voice traffic. In a data exchange, typically more data is sent in one direction and less in the other. For instance, if fax data were being sent to a MS, a higher data rate would be advantageous in the BS to MS direction. In this case the majority of the TDD slot time is given to the MS to BS link. A particular BS to MS and/or MS to BS may also be assigned multiple slots within a single polling loop. This can dramatically increase the bearer channel data every second or every fourth polling loop, the data rate is cut significantly. This technique is useful for instrument telemetry data, fax and video control back channel, etc.

While a preferred embodiment is disclosed herein, many variations are possible which remain within the scope of the invention as defined in the appended claim, and these variations would become clear to one of ordinary skill in the art after perusal of the specification, drawings and claim herein.

For example, information which is transmitted from transmitter to receive is referred to herein as "data", but it would be clear to those of ordinary skill in the art, after perusal of this application, that these data could comprise data, voice (encoded digitally or otherwise) error-correcting codes, control information, or other signals, and that this would be within the scope of the invention.

Moreover, while the specification has been described with reference to TDMA multiplexing of air channels, it would be clear to those of ordinary skill in the art, after perusal of this application, that air channels may be multiplexed by other means, including FDMA (frequency division multiplexing), by assigning air channels to differing frequency bands, CDMA (code division multiplexing), by assigning air channels to differing spread-spectrum spreading codes, other multiplexing techniques, or combinations of these multiplexing techniques, and that this would be within the scope of the invention.

Claims

A method of establishing and maintaining communication between a base station (104) and one of a plurality of user stations (102) in a time division multiplexed communication system, the method comprising the steps of:
transmitting a general poll message from said base station in a currently unoccupied slot;

receiving said general poll message at said one user station;

transmitting a general poll response from said user station, said general poll response including an identification of said user station;

receiving said general poll response at said base station;

transmitting a first specific poll message from said base station inviting said user station to seize the current slot;

receiving said first specific poll message at said one user station;

transmitting a first specific poll response from said one user station;

receiving said first specific poll response at said base station;

thereafter transmitting and receiving information messages between said base station and said one user station over an established communication link;

measuring at said user station the quality of said communication link between said base station and said one user station and of other free time slots with said base station;

transmitting if the measured quality of said communication link is below a threshold, a hand off message request from said user station;

receiving said hand off message request at said base station;

transmitting a second specific poll message from said base station;

receiving said second specific poll message at said one user station;

transmitting a second specific poll response from said one user station in response to said second specific polling message;

receiving said second specific poll response at said base station; and

thereafter transmitting and receiving further information messages between said base station and said one user station over a re-established communication link.