JPH03217132A - Time division communication method and system for mobile communication - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a time-division communication method for a wireless communication channel in mobile communication and the effective use of multiple load gain of a time compression multiplexed signal, which is a modulation signal, in the system. More specifically, a radio channel is given, and one of the many mobile radios in the service/factory uses this to set up a radio link and communicate with the opposing radio base station. When another mobile radio starts communicating with another radio base station using the same radio channel, due to the effective use of frequencies or radio wave propagation characteristics, An object of the present invention is to provide a method for eliminating adverse effects on communication with a wireless base station and at the same time improving the effective use of frequencies by reducing transmission output, and an economical system using the method. It is.

[Prior Art] In mobile communication using voice using a small zone method, a method employing a time division time compression multiplex signal is described in the following document.

Literature 1. Ito ``Mobile phone system study Temporary division time compression F
Proposal of M modulation method 1'' IEICE technical report RCS89-11
July 1989 document 2. Ito'゛Mobile phone system study: theoretical study of temporary division time compression FM modulation system'' IEICE technical report RCS89-3
9 October 1989 3 In other words, in Document 1, a transmission signal (Bestband signal) is divided into predetermined time intervals and stored in a memory circuit, and when read out, the speed is n times faster than the speed at which it is stored in the memory circuit. built into mobile radios and radio base stations in order to read out signals in predetermined time slots at high speed, angle-modulate or amplitude-modulate carrier waves with the signals accommodated by these time slots, and transmit and receive signals intermittently in time. a radio receiving circuit having receiving mixers facing each other and communicating with each other, a radio transmitting circuit having a transmitting mixer, a synthesizer applying voltage to the receiving mixer of the wireless receiving circuit, and a synthesizer applying voltage to the transmitting mixer of the wireless transmitting circuit. A switch circuit is provided for each to intermittent the output of the synthesizer applied to each, and this intermittent state is synchronized for both transmission and reception, and the same intermittent transmission and reception as above is synchronized for the radio base station communicating with the mobile radio. In order to extract only the signals accommodated in the predetermined time slots, the receiving side opens and closes the radio receiving circuit to receive the signals, demodulates the signals, and stores the obtained signals in the storage circuit. An example of a system has been reported in which a system is constructed that makes it possible to reproduce the baseband signal, which is the original signal that has been transmitted, by reading it out at a low speed that is 1/n of the speed at which it is stored in this memory circuit. has been done.

In addition, Reference 2 examines adjacent channel interference and co-channel interference, which are problems when the above-mentioned CM (time division time compression multiplexing)-FM method is applied to small zones, and the system・The possibility of realizing the system has been shown by appropriately selecting parameters.

Furthermore, the multiple load gain of a so-called frequency division multiplexed signal, which is obtained by converting the frequency of an audio signal and multiplexing it so that it does not overlap on the frequency axis, is described in, for example, the following documents 3 and 4.
is shown.

Reference 3.8, D. Holbrook, J. T. Di
xon: LoadRating Theory for
r Multichannel Amplifiers
, BSTJ, 1B, Oct. , 1939 Reference 4
．． C.B. Feldman et al.”Band Width
andTransmission Performan
ce”BSTJ.July 1949490～
Fig. 8 on page 595 is the same as Fig. 8 of the above-mentioned document 3. 7, and Figure 9 is quoted from page 495 of the above-mentioned document 4, and shows that it is possible to obtain substantially the same multiple load gain as that shown in Figure 8. ing.

The reason why multiple load gain can be obtained and its application to wireless angle modulation will be briefly explained below.

The level of a telephone channel in operation, through which telephone signals are flowing, varies depending on the person, gender, and subscriber line length, and even when the same person is speaking continuously, there is a There is always a gap. Also, while the other party is talking, the other party does not speak and no signal is added to one direction. Do not speak during exchange connection. Therefore, the individual signal levels are diverse, and a composite signal of these cannot be easily obtained. However, clarifying this is the most important and fundamental problem in creating a relay line that maintains distortion, crosstalk, quasi-crosstalk, noise, etc. to satisfactory values. Therefore, it has been studied by many people.

FDM method (SS: SinCIIe) with carrier wave suppression
The level of "method applying sideband" is to synthesize such voices, and the probability that each voice overlaps at the same time is rare, and while the number of communication paths N is small, each voice that fluctuates greatly has no direct effect on the synthesized signal. However, as the number of multiplexes increases, the individual effects become less direct and become stochastically averaged out. Therefore, the peak value of the composite signal increases very slowly as the number of communication paths increases. This is B. D. Hotbrook and J. T. Dixo
n (Reference 3 above) was statistically determined for US telephones. According to the results, the change in power of the limited wave having the same peak value as the peak value of the multiplexed signal is as shown in FIG. To show how small the increase in the peak voltage of a multiplex telephone signal is, when compared with the sum of the peak voltages of the individual signals, "
It becomes like a multi-load gain. That is, for example 960
In the communication path method, six communication paths are loaded at the maximum at the same time, and the peak voltage is the same as if the signal of 954 communication paths were not loaded.

7 In the SS-FM system, fluctuations in the voltage of the composite signal result in frequency deviations, so when the peak frequency deviation of the composite signal is set to a certain value, and the number N of multiplexed communication paths increases, each communication signal becomes a voltage sum. The modulation index per channel can be increased by the multiple load gain shown in Figure 8 compared to when Improved.

[Problem to be solved by the invention] In the system construction examples shown in the above-mentioned documents 1 and 2,
The existence of a multiple load gain in time-division time compression multiplexed signals transmitted from a radio base station to a large number of mobile radios is not disclosed, and this multiple load gain is not utilized.

Therefore, if this multiple load gain analysis had been performed, many advantages would be gained in system design, such as the reduction in transmit power level that is possible by increasing the depth of frequency modulation. , Ding C
Specific advantages include ease of designing an amplifier for amplifying the M signal, realization of an economical amplifier by expanding the operating level setting range, and economy by relaxing the rating conditions of mixers, resistors, and capacitors. There was a problem that needed to be solved that could not be done.

What is disclosed in References 3 and 4 is to clarify the multiple load gain in so-called frequency division multiplexed signals, in which audio signals are frequency-converted and multiplexed so that they do not overlap on the frequency axis, and are time-division time multiplexed. It cannot be applied to compressed multiplexing (TCM) signals, and the existence of multiple load gain is unknown. Advantages (References 1 and 2 above)
There was a problem that needed to be solved, namely that it was not possible to concretely realize the same thing in the case of .

Means for solving the L problem Kocho CM (time division time compression multiplexing) Taking the multiplex number (number of communication paths) of the signal, the time length of one frame, and the highest frequency of the original signal as parameters, Using the sampling theorem, the load gain was clearly derived in relation to the multiple load gain in FDM (frequency division multiplexing@), and this was made practical.

[Effect] Since it has become clear that multiple load gain exists in TCM signals, it is now possible to specifically calculate multiple load gain using various design parameters of the system, and it is possible to reduce interference, etc. to acceptable values. FM (PM
) By deepening the degree of modulation, it was possible to gradually reduce the transmission output. Therefore, the amplifier design becomes easy and
Additionally, the rated values of passive circuits such as mixers, resistors, and capacitors can be lowered, making it possible to construct an economical system.

[Embodiment] FIG. 1A, FIG. 1B, and FIG. 1C show a system configuration for explaining an embodiment of the present invention.

In FIG. 1A, 10 is a general telephone network, and 20 is a gateway exchange for connecting the telephone network 10 and the wireless system. 30 is a wireless base station and gateway switch 2
0, a circuit for converting signal speeds, a circuit for allocating and selecting time slots, a control unit, etc. 100-n) and a wireless transmitting/receiving circuit for transmitting and receiving wireless signals.

Here, between the barrier switch 20 and the wireless base station 30,
There are transmission lines for transmitting communication signals 22-1 to 22n including communication signals of communication channels CI-11 to CI-In and control signals.

FIG. 1B shows a circuit configuration of a mobile radio device 100 that communicates with a radio base station 30.

Received signals such as control signals and call signals received by the antenna section enter a radio receiving circuit 135 that includes a receiving mixer 136 and a receiving section 137, and the output communication signal is sent to a speed restoration circuit 138, a control section 140, and a clock signal. The signal is input to the regenerator 141. The clock regenerator 141 regenerates a clock from the received signal and applies it to the speed recovery circuit 138, the control section 11140, and the timing generator 142.

The speed recovery circuit 138 calculates the speed (pitch in the case of an analog signal) of the compressed and segmented communication signal in the received signal.
is restored and input as a continuous signal to telephone section 101 and control section 140.

The communication signal output from the telephone unit 101 is divided into predetermined time intervals by a speed conversion circuit 131, and the speed (pitch in the case of an analog signal) is made high (compressed) and sent to a transmission mixer 133. The signal is applied to the wireless transmission circuit 132 including the section 134.

The output of the modulator included in the transmitting section 134 is transmitted to the transmitting mixer 13
3, the signal is converted to a predetermined radio frequency, transmitted from the antenna section, and received by the radio base station 30.

In order to send a radio signal from the mobile radio device 100 to the radio base station 30 using a time slot that is permitted to be used, timing information from the timing generator 142 shown in FIG. One thing that is necessary is to be able to obtain it through

In this timing generator 142, the clock regenerator 14
1 and a control signal from the control section 140, necessary timing is supplied to the transmission/reception intermittent controller 123, speed conversion circuit 131, and speed restoration circuit 138.

This mobile radio device 100 further includes a synthesizer 121.
-1 and 121-2, and selector switch 122-1.1
22-2, a transmission/reception intermittent controller 123 and a timing generator 142 that generate signals for switching the changeover switches 122-1 and 122-2, respectively.
Synthesizers 121 - 1 and 121 - 2 , transmission/reception intermittent controller 123 , and timing generator 142 are controlled by control section 140 . Each synthesizer 121-1.1
21-2 is supplied with a reference frequency from a reference crystal oscillator 120.

A wireless base station 30 is shown in FIG. 1C.

N-channel communication signal 22-1 with gateway switch 20
22-n are connected to a signal processing section 31 forming an interface through a transmission line.

Now, the communication signal 22- sent from the barrier switch 20
1 to 22-n are input to the signal processing unit 31 of the wireless base station 30. The signal processing section 31 is equipped with an amplifier for compensating for transmission loss, and also performs so-called 2-wire to 4-wire conversion. That is, the input signal and the output signal are mixed and separated, and the input signal from the barrier switch 20 is sent to the signal speed conversion circuit group 51. Also, signal speed restoration circuit group 3
The output signal from 8 is transmitted to the barrier exchange 20 by the signal processing unit 31 using the same transmission path as the input signal. Among the above input signals from the barrier switch 20, the signal speed conversion circuit group 5 including many signal speed conversion circuits 51-1 to 51-n
1 and undergoes speed (pitch) conversion at predetermined time intervals.

In addition, the signal transmitted from the wireless base station 30 to the barrier exchange@20 is such that the output of the wireless receiving circuit 35 is transmitted to the signal selection circuit group 39.
is input to the signal speed restoration circuit group 38 via the signal speed (
(pitch) is converted and input to the signal @ processing section 31.

Now, the control of the radio reception circuit 35 or the output of the call signal is performed by a signal selection circuit 39 that selects a signal for each time slot.
-1 to 39-n are input to the signal selection circuit group 39,
Here, speech signals are separated corresponding to each speech channel CI-11 to CHn. After this output undergoes signal speed (pitch) restoration in a signal speed restoration circuit group 38 including signal speed restoration circuits 38-1 to 38-n provided for each channel, it is input to the signal processing section 31. , after undergoing 4-wire to 2-wire conversion, this output is sent as a communication signal 22-1 to the barrier switch 20.
~22-n.

Next, the functions of the signal speed conversion circuit group 51 will be explained.

By storing input signals such as audio signals and control signals divided into a certain length of time in a storage circuit, and changing the speed when reading them out, for example, reading them out at a speed 15 times faster than when they were stored, the signal can be read out. It becomes possible to compress the time length. The principles of the five signal speed conversion circuit groups are the same as when playing back audio recorded by a tape recorder at high speed.1. 5, and in fact, for example, C CD (Char
geCoupledDevice). BBD
(Bucket Brigade Device) can be used, and the memory used in television receivers and tape recorders that compress or expand the time axis of conversations can be used (References: Kosaka et al.
``A tape recorder that compresses/expands the time axis of conversations''
Nikkei Electronics July 26, 1976 92-
Page 133》.

The circuit using COD or BBD exemplified in the signal speed conversion circuit group 51 can be used as it is in the signal speed restoration circuit group 38 as described in the above-mentioned literature. This can be achieved by making the reading speed slower than the writing speed by receiving a timing signal from a timing generator 42 that generates timing based on the clock and a control signal from the control unit 40.

Control or audio signals outputted from the barrier switch 20 via the signal processing section 31 are input to the signal speed conversion circuit group 51, and after speed (pitch) conversion processing is performed, the signals are converted for each time slot. is applied to the signal assignment circuit group 52 that assigns the . The signal allocation circuit group 52 is a buffer memory circuit that stores one section of high-speed signals output from the signal speed conversion circuit group 51, and receives timing information from the timing generation circuit 42 given by instructions from the control section 40. Then, the signal in the buffer 7 memory is read out and transmitted to the wireless transmission circuit 32. As a result, when viewed in terms of channels, communication signals are chronologically arranged in series without overlapping, and when all control signals or communication signals, which will be described later, are not implemented, they appear as if they were continuous signal waves. become.

The state of this compressed signal is shown and explained in FIGS. 2A and 2B.

The output signal of the signal speed conversion circuit group 51 is sent to the signal assignment circuit group 5.
2 and are given time slots in a predetermined order. Figure 2A (a> SDI.SD2
−． SDn indicates that the speed-converted communication signals are allocated to each time slot.

Here, one time slot accommodates a synchronization signal and a control signal or a call signal as shown in the figure. If the call signal is not implemented, only the synchronization signal is used and an empty slot signal is added to the call signal portion. In this way, as shown in FIG. 2A(a), the wireless transmitting circuit 32
In this case, signals forming one frame are applied to the modulation circuit in time slots SD1 to SDrl.

The time-series multiplexed signal to be transmitted is angularly modulated in the radio transmission circuit 32, and then sent out into space from the antenna section.

When making or receiving a telephone call, the wireless base station 30
Regarding the transmission of control signals between the mobile radio device 100 and the mobile radio device 100, it is possible to use either within the telephone signal band or outside the telephone signal band.

Figure 3A shows these frequency relationships. In other words, in the figure (a), it is an example of an out-of-band signal, and as shown in the figure,
Low frequency side (2501-1z) and high frequency side (3850Hz
) can be used. This signal is used, for example, when it is desired to send a control signal during a call.

FIG. 3A (b) shows an example of an in-band signal, which is used when making and receiving calls.

Although the above examples were all tone signals, it is possible to transmit many types of signals at high speed by increasing the number of tone signals or by modulating the tone and making it into a subcarrier signal.

The above was a case of analog signals, but if a digital data signal is used as a control signal, it is also possible to digitally encode the audio signal and transmit the two by time division multiplexing. The circuit configuration of is shown in FIG. 3C. FIG. 3C shows an audio signal digital encoding circuit 9.
This is an example of a case in which the signal is digitized at 1, multiplexed with a data signal by a multiplex conversion circuit 92, and applied to a modulation circuit included in the wireless transmission circuit 32. However, in digital data signals, there is usually no multiple load gain when multiplexing analog signals, which will be described later, so this point must be kept in mind when designing the system.

19 Then, it is received by the opposite receiver, and the 3rd C is received by the demodulation circuit.
By performing the operation opposite to that shown in the figure, it is possible to extract the audio signal and the control signal separately.

On the other hand, the signal sent from the mobile radio device 100 is received by the antenna section of the radio base station 30 and input to the radio reception circuit 35. FIG. 2A (b) schematically shows this upstream input signal.

That is, time slots SU1, SU2 . ...S
Un is the mobile radio device 100-1, 100-2. ...,
100-n shows a transmission signal addressed to the wireless base station 30.

Also, each time slot su”t, su2,...,s
If the contents of un are shown in detail, as shown in the lower left of FIG. 2A (b), it consists of a synchronization signal and a control signal or/and a telephone call signal.

However, if the distance between the radio base station 30 and the mobile radio device 100 is small or depending on the signal speed, it is possible to omit the synchronization signal. Furthermore, the waveform of the radio carrier wave of the above-mentioned uplink radio signal within a time slot is schematically shown as shown in FIG. 2B (C>20).

Now, among the input signals that have arrived at the wireless base station 30, the control signal is immediately sent to the control unit 40 from the wireless receiving circuit 35.
added to. However, depending on the magnitude of the speed conversion rate, it is also possible to apply the output of the signal speed restoration circuit group 38 to the control unit 40 after performing similar processing on the call signal. Further, the call signal is applied to the signal selection circuit group 39.

A timing signal from a timing generation circuit 42 that generates a predetermined timing is applied to the signal selection circuit group 39 in response to a control signal instruction from the control unit 40, and a synchronization signal and a control signal are generated for each time slot SU1 to Sun. The signal or speech signal is separated and output. Each of these signals is input to a signal speed restoration circuit group 38. This circuit is a speed conversion circuit 131 (first
It has a function to perform the inverse conversion shown in Figure B), thereby faithfully reproducing the original signal and transmitting it to the gateway exchange 20.

The manner in which signals are transmitted in the signal space according to the present invention will be explained below using the required transmission band and the relationship between this and adjacent wireless channels.

As shown in FIG. 1C, the control signal from the control section 40 is applied to the wireless transmission circuit 32 in parallel with the output of the signal allocation circuit group 52. However, depending on the magnitude of the speed conversion rate, it is also possible to apply the signal to the wireless transmission circuit 32 from the output of the signal allocation circuit group 52 after performing the same processing as the call signal.

Next, as shown in FIG. 1B, the mobile radio device 100 also has a circuit configuration that is required when the radio base station 30 functions as one communication channel.

Original signal such as audio signal (0.3KHz~3.0KH
z) passes through the signal speed conversion circuit group 51 (FIG. 1C), the frequency distribution on the output side is as shown in FIG. 3B. In other words, if the audio signal is converted 15 times as described above, the frequency distribution of the signal is as shown in Figure 3B.
This means that it has been expanded to 4.5KHZ to 45KHZ. Here, the frequency distribution of the signal is expanded, but the waveform form is simply stretched along the frequency axis;
It should be noted that the waveform itself does not change.

This is necessary when determining the value of multiple load gain. Now, FIG. 3B shows a case where the control signal is simultaneously transmitted using the lower frequency band of the audio signal.

Among these signals, the control signal (0.2 to 4.0KHZ) and the call signal CH1 (SD at 4.5 to 45KHZ)
(denoted as I) is accommodated in a time slot, say SD1.

The same applies to the audio signals accommodated in the other time slots SD2 to SDrl.

That is, time slot SDi (i=2.3,
・．． n> is a control signal (0.2 to 4.OK+-1
2 > and communication signals Ql-1i (4.5 to 45 KHz) are accommodated. However, the signals in each time slot are arranged in chronological order, and the signals in multiple time slots are are not applied to the wireless transmission circuit 32 at the same time.

These call signals are sent to the wireless transmission circuit 32 along with control signals.
When added to the angle modulation section included in the above, the required transmission band requires at least 23 fo ±45 KHZ. However, fo is a radio carrier frequency. If there are a plurality of wireless channels given to the system, the speed-up of the signal by the signal speed conversion circuit group 51 is limited to a certain value due to the limitations of these frequency intervals. The frequency interval of multiple wireless channels is f
,. , and the maximum signal speed due to the above-mentioned speed increase of the audio signal is fH, then the following inequality must hold between the two.

f > 2f ■ rep On the other hand, in digital signals, audio is usually digitized at a speed of about 54 kb/s, so it is necessary to raise the scale on the horizontal axis by about one digit in Figure 3B, which explains the case of analog signals. However, the relationship in the above equation also holds true in this case.

Further, the control signal input from the mobile radio device 100 to the radio base station 30 is input to the radio reception circuit 35, but a part of its output is input to the control unit 40, and the other part is input to the signal selection circuit group 39. The signal is then sent to the signal speed restoration circuit 24 and the path group 3B. After the latter control signal undergoes speed conversion (conversion to a low speed signal) that is completely opposite to that at the time of transmission, it becomes the same signal speed as that used in the general telephone network 10 and is transmitted via the signal processing section 31. Sent to barrier exchange 1120.

Next, the call originating/receiving operation of the system according to the present invention will be explained by taking the case of a voice signal as an example.

(1) Call origination from mobile radio device 100 This will be explained using the flowcharts shown in FIGS. 4A and 4B.

When the mobile radio device 100 is powered on, the first
In the wireless receiving circuit 135 in Figure B, the downlink (wireless base station 30
→Mobile radio device 100) Starts capturing the control signal included in the radio channel (referred to as channel CHI). If the system is provided with multiple radio channels, i》 Radio channel 11 exhibiting the highest received input electric field;
》 Wireless channels instructed by control signals included in the wireless channel ii 》 Channels with empty time slots among the time slots in the wireless channel, etc. The reception state of the channel Cl {1} is entered. This is possible by capturing the synchronization signal within the time slot SDi shown in FIG. 2A (a). 121
-1 to generate a local frequency that enables reception of the radio channel CH1, and the switch 122-1 is also tilted and fixed to the synthesizer 121-1 side.

Therefore, when the receiver of the telephone unit 101 is off-hook (starts making a call) (S201, FIG. 4A), the synthesizer 121-2 of FIG. 1B generates a local frequency that enables transmission of the wireless channel CH1. A control signal is received from the control section 140 to cause the control to occur.

Further, the switch 122-2 is also turned toward the synthesizer 121-2 side and becomes fixed. Next, wireless channel CH1
The call control signal outputted from the telephone unit 101 is sent using the telephone unit 101. This control signal has a frequency band shown in FIG.
It is sent using un.

The transmission of this control signal is limited to time slot Sun, and is sent in bursts, and no signal is transmitted during other time slots, so it does not adversely affect other communications. However, if the speed of the control signal is relatively slow or the amount of information in the signal is large and cannot be accommodated in one time slot, the same time slot of the next frame or Sent using slots.

Specifically, the following method is used to capture the time slot Sun. As shown in FIG. 2A (a), the control signal transmitted from the radio base station 30 includes a synchronization signal and a subsequent control signal, and the mobile radio device 100 receives this signal to perform frame synchronization. becomes possible.

Roughly speaking, this control signal includes control information such as currently used time slots and unused time slots (empty time slot display). In some systems, time slotsor <i=1.2. ..., n) may contain only synchronization signals and call signals when they are being used by other communications, but even in such cases, unused time slots usually contain synchronization signals and control signals. By receiving this control signal, the mobile radio 11100 can know which time slot to use to send out the calling signal.

Note that if all time slots are in use,
It is not possible to make a call on this radio channel and it is necessary to sweep and search for another radio channel.

In other systems, there may be no empty slot indication in any of the time slots, in which case the following audio multiplex signals SDI, SD2 . ...,S
It is necessary to search for the presence or absence of Dn one after another and check the empty time slots 1 to 1.

Now, returning to the main topic, the mobile radio a 100, which has received the control information sent from the radio base station 30 using one of the above methods, determines in which time slot it should send the call control signal. The determination can be made including the transmission timing.

Therefore, assuming that the time slot Sun for uplink signals is an empty slot, it is decided to use this empty time slot, and the necessary timing is determined from the response signal from the radio base station 30 by sending out a control signal for calling. burst-like control signals can be sent out.

If there is a call from another mobile radio at the same time, the calling signal will not be properly transmitted to the radio base station 30 due to call collision, and the operation will have to be restarted from the beginning, but this probability is When viewed as a system, this value is kept to a sufficiently small value. If call collisions are to be further reduced, the following method may be used. If the mobile radio device 100 finds an empty time slot in which it can make a call, it does not use the entire time slot, but rather uses only the first half of the time slot for some mobile radio devices and only the second half for other mobile radio devices. This is a method that allows you to use it. In other words, the time signal is used as a calling signal.
Divide the used portion of the slot into several types, use this to classify a large number of mobile radio devices into groups, and place one of the slots in each group.
This method gives the time period within one time slot.

Another method is to create many different frequencies for the control signal,
This is a method of dividing a large number of mobile wireless devices into groups and applying this to each group. According to this method, even if control signals of different frequencies are transmitted simultaneously using the same time slot, no interference will occur at the radio base station 30. The above two methods may be used separately, or when used together, the effects will increase synergistically.

Now, suppose that the call control signal from the mobile radio device 100 is successfully received by the radio base station 30 and the ID (identification number) of the mobile radio device 100 is detected (8202), the control unit 40 determines that the ID (identification number) of the mobile radio device 100 is currently vacant. Search for a time slot.The time slot given to mobile radio 100 may be Sun, but just to be sure, perform a search.
In addition to 00, this is also used to respond to simultaneous calls from other mobile radios, and to provide time slots suitable for service types and service classifications.

As a result, for example, if the time slot SDI is vacant, the time slot SDI of the radio channel CH1 is used for the mobile radio 11100 to send the time slot upstream (from the mobile radio 100 to the radio base station 30) by a downlink control signal. It instructs to use SU1 and the corresponding downlink (radio base station 30→mobile radio device 100) SDI (3203). Accordingly, the mobile radio device 10
0, the state transitions to a state where reception is possible at the designated time slot SDI, and the downstream time slot SD
SU1 (see FIG. 2A (b)), which is the time slot for the uplink radio channel corresponding to I, is selected. At this time, the control unit 140 of the mobile radio@100 operates the transmission/reception intermittent controller 123, The switches 122-1 and 1222 are started operating (3204>. At the same time, a slot switching completion report is sent to the upstream time slot SU1.
(8205) 31 and waits for a dial tone (S206>).

Time slot SU of the radio carrier of this upstream radio signal
The state of No. 1 is schematically shown in FIG. 2B (C). In addition to the time slot SU1, the radio base station 30 receives an uplink signal SU3 from another mobile radio device 100.
and Sun are included in one frame and sent.

The radio base station 30 that has received the slot switching completion report (
3207), and sends a calling signal to the gateway exchange 20 (
8208), and upon receiving this, the gateway exchange 20 detects the ID of the mobile radio 100, and turns on the necessary switch among the switch group included in the barrier exchange 20 (S20
9>, send a dial tone (S210, 4th B)
figure>.

This dial tone is transferred by the wireless base station 30 (3211>, and the mobile wireless device 100 confirms that the communication path has been set (S212>).

When transitioning to this state, the telephone unit 1 of the mobile radio device 100
Since a dial tone is heard from the 01 handset, the dial signal begins to be sent. This dial signal 32 is speed-converted by a speed conversion circuit 131 and is converted into a wireless transmission circuit 1 including a transmission section 134 and a transmission mixer 133.
32 using the uplink time slot SU1 (3213>. Thus, the transmitted dial signal is received by the radio receiving circuit 35 of the radio base station 30. In this radio base station 30, the mobile radio 100, gives the time slot to be used, operates the signal selection circuit group 39 and the signal allocation circuit group 52 of the radio base station 30, and receives the uplink time slot SU1. The dial signal transmitted from the mobile radio 11100 passes through the signal selection circuit 39-1 of the signal selection circuit group 39. The original transmission signal is input to the signal speed restoration circuit group 38, where the original transmission signal is restored, and is transferred to the barrier exchange 1fi 20 as a call signal 22-1 via the signal processing unit 31 (3214>, and the call path to the telephone network 10 is Set (8215>).

On the other hand, an input signal from the barrier switch 20 (initial control signal,
When a call is started, the call signal) undergoes speed conversion at the signal speed conversion circuit group 51 in the radio base station 30, and is given a time slot SDI by the signal allocation circuit 52-1 of the signal allocation circuit group 52. ing. Then, the time slot SD of the wireless channel downstream from the wireless transmission circuit 32
1 to the mobile radio device 100. In the mobile radio device 100, the time and time of radio channel CH1 are
It is waiting for reception in the slot SDI, is not received by the radio receiving circuit 135, and its output is input to the speed restoration circuit 138. In this circuit, the original signal of the transmission is restored,
It is input to the handset of the telephone unit 101. Thus, a call is started between the mobile radio device 100 and a regular telephone within the regular telephone network 10 (8216>).

The call is terminated by on-hooking the handset of the telephone unit 101 of the mobile radio device 100 (S217>, the end-of-call signal and the on-hook signal from the control unit 140 are transmitted to the speed conversion circuit 1.
31 from the wireless transmission circuit 132 to the wireless base 830 (S218>, the control unit 140 stops the operation of the transmission/reception intermittent controller 123, and switches 1221 and 122-2 to the synthesizer 1
It is fixed to the output ends of 21-1 and 121-2.

On the other hand, in the control unit 40 of the radio base station 30, the mobile radio device 1
When the call termination signal from 00 is received, the call termination signal is transferred to the barrier switch 20 (S219>, and the switch group (not shown)
The switch is turned off to end the call (S220>. At the same time, the signal selection circuit group 39 and signal allocation circuit group 52 in the wireless base station 30 are opened.

In the above explanation, the exchange of control signals between the radio base station 30 and the mobile radio device 100 was explained as not passing through the signal speed conversion circuit group 51, the signal speed restoration circuit group 38, etc.
This is for convenience of explanation, and communication can be carried out without any problem even through the signal speed conversion circuit group 51, signal speed restoration circuit group 38, control signal speed conversion circuit 48, or signal processing section 31, as with audio signals. .

35 (2) Incoming call to mobile radio device 100 The mobile radio device 100 is on standby with the power turned on. In this case, as explained in the section regarding the call origination from the mobile radio 1100, the mobile radio 1100 is in a waiting state to receive a downlink control signal of the radio channel CH1 according to the procedure defined by the system.

Assume that an incoming call signal to the mobile radio device 100 arrives at the radio base station 30 from the general telephone network 10 via the barrier switch 20. These control signals are transmitted as communication signals 22 to the control unit 40 (FIG. 1C) via the signal rate conversion circuit group 51 and the signal allocation circuit group 52, similarly to the voice signals. Then, the control unit 40 uses an empty slot among the downlink time slots of the radio channel C1''1 addressed to the mobile radio device 100, for example, SD1, to
Signal 10 Incoming call signal display Signalman time slot usage signal (
For example, SU1 corresponding to SDI is used for transmission from the mobile radio device 100.

In the mobile radio device 100 that receives this signal, it is transmitted from the receiving section 137 of the radio receiving circuit 135 to the control section 1409 6 . The control unit 140 confirms that this signal is an incoming call signal to its own mobile radio 100, so it makes a ring tone from the telephone unit 101 and at the same time waits at the designated time slots SD1 and SU1. The transmission/reception intermittent controller 123 is operated as shown in FIG. In this way, the state has shifted to a state in which calls can be made.

It should be noted that it is theoretically explained in Reference 2 that signal transmission is performed in good condition using this system and does not adversely affect other wireless channels in the system, so it will be omitted here. It will be theoretically explained that the TCM signal applied to the present invention has multiple load gain, and then its application will be described.

(3) Regarding the multiple load gain of the TCM signal transmitted from the radio base station 30 In order to correlate the multiple load gain of the TCM (time division time compression multiplexing) signal with the multiple load gain of the FDM (frequency division multiplexing) signal, first, , each channel CH1 of FDM
, CH2. ...Each audio signal flowing through CHn is expressed in the form of a function. As is well known, FDM signals are frequency-converted audio signals and arranged in a line on the frequency axis as shown in Figure 5. This multiplexed signal is often used in coaxial transmission systems and microwave analog communication systems. , multiple load gain has also been incorporated into practical systems and is showing great effects. Although the spectrum in FIG. 5 shows the case of 12 channels (CH1 to 12), in general, in addition to 12 channels, 60, 120, 480, 960, 1200.27
Many different types are used, such as 00 pieces. Now, the fifth
Channels CH1, CH2, CH3, --...- in the figure. C
Audio signal flowing to t-In (In the case of wired, the frequency band to be transmitted is 0.3 to 3.4 KHZ, but in the case of mobile radio signals, it is 0.3 to 3.0 KHZ, so it was limited to this value) f1m, f2(o,..., fn(t)
shall be.

The frequency components of these signals are 0.3 to 3.0 KHz for channel CH1 and 4.3 to 7.0 KH for CH2.
z, ---・-. CHn is 4x (n-1>10.3 ~
4X (n-1 > KHz, so they do not overlap with each other. However, when viewed from the signal waveform, the amplitude distribution of these n audio signals is simply shifted to a higher frequency on the frequency axis. The signal waveform itself does not change at all.This is important in determining the multiple load gain, and can be expressed as follows.

The known multiload gain of the FDM signal is exactly the same as the signal when n voices are arranged on the frequency axis as shown in FIG. 5, and the signal when n voices are simply mixed without frequency conversion. Prove this using a formula. Channel CH1,C}-f
2. ..., the mixed signal of CI-1rl is expressed by the following equation.

F(t)=f1(t)+f2(t) 10-+f,(t)
(1) Specifically, fi(t) can be expressed as follows.

3q 3kHz (2) (3) However, 1≧2 Also, the mixed signal when frequency conversion is not performed is given by the following equation.

G (t) -g1 (1) + Q 2 (1) Ten...+Qo (1) (4) Here, g1 (t)=f1 (1) (5) (6) however, i≧2 next, (1), Power possessed by the signal of the formula 《4》 seek.

first, F (t) The power of +f2 (1) 2 Ten...+f, (1) 2 (7) on the other hand, G(t) The power of 0 + Q 2 (1) ? Ten... ten go (1) 2 (8) however, No power is formed between different signals. I used a word.

That is, τ 41 →　υ τ f Cli Qj dt=0 0 (9) However, i≠j Now, from (2>.(3) equation), (10) (11) However, 1≧2 From equations (10) and (11), F(t)2=G(t)2(12) is obtained.

That is, it has become clear from the above explanation that the power of a signal is not related to frequency conversion.

42 Next, consider applying the sampling theorem to Q+(t). Since the highest frequency fh of J(j) is 3KHz, the time interval is 1/(2fh), that is, 1〆600
It is well known that if sampling is performed every 0 seconds, the original signal can be reproduced later even if only the sample value (voltage value) is transmitted.

Therefore, as shown in Figure 6A (a), f, (1) is calculated using the time interval [1/6000+ (1/6000)( i-1)/
6000] seconds (13). In the same figure, t, =
1/6000 second, t b= (1/6000) X (1/6000
) seconds. tC= (1/6000)
/6000) seconds, t6 =1/6000+(1/
6000) x (3000 /6000) seconds,
t o = 1/6000 seconds. Hereinafter, in order to explain specifically, the multiplex number n is set to 6.
000, 'l The frame length is 1/6000 seconds.

Now, on the horizontal axis of Figure 6A (a), put 600 small bags (diameter 1/6000 x 1/6000 seconds) as shown in Figure 6B (C).
0 pieces, 43 One large bag with a diameter of 1/6000 seconds will be arranged as shown in the figure. Then, compare the sampling values above to matchsticks and consider how these matchsticks fit into each bag.

The function QJ m (i=1.2, -, n) has 6000 matchsticks per second, and they are equally spaced in time, so each bag (1), (2), ..., (N)
One line is sent to each frame, and this operation is repeated for each frame. In other words, Bag (1) has Q1 (1/6000) Bag (2) has g2 (1/6000+1/600
0x 1/6000) bag (3) contains g3 (1/
6000+1/6000X 2/6000) bag (600
0) will be filled with q(1/6000+1/60006000 x5999/6000), respectively.

Also, in the large bag (Σ6000), the mixed signal G(t
) is included. In this case, the sampling time is 1/600044 + 1/6000X 3000/6000, that is, the midpoint of one frame. Then, a large bag (Σ600
0 ) will be filled with the value of the next matchstick.

The large bag (Σ6000) has G (1/6000 + 1/6000x 3000
/6000) or less, the value of the matsushi stick from bag (1) to (6000), g1 (1/6000) Q2 (1/6000+ 1/6000X 1/60
00) ,Q 6000 (1/6000+1/60
00X 5999 /6000) and the large bag (Σ60
00) and the value of the matchstick inserted in G (1/6000 + 1/6000x3000 /6
000) below (14). Compare as shown in equation (15).

+Q2 (1/6000+1/6000xl/600
0) 10 Cl 3 (1/6000+1/6000x 2
/6000)45 +gH (1/6000+1/6000x(i-1)
/6000) 10Q 6000 (1/6000+1
/6000x 5999 /6000) (14) G (1/6000+1/6000x3000/600
0 ) = Q 1 ( 1/8000+ 1/800
0X 3000/6000 ) 102 (1/600
0+1/6000x3000/6000)+・・・
・・+Qo(1/6000+1/6000x3000/60
00 ) (15) As a result of comparing equations (14) and (15), if 6000 18 (16) and Δ is sampled at an interval of 1/6000 seconds, its average value converges to O. This means that it has been proven that the multiload gain in the FDM signal is similar and has the same value in the TCM signal as well. Because 6000 placed on the horizontal axis
bags, each bag representing one time slot, the sum of the bags being one frame, and the matchsticks in the bags each signal c+, c+ . ．． g is hour and minute 1 2 ゜”
'' It can be thought of as a signal subjected to time compression multiplexing (CM>), and T
This is because the CM signal is well mixed, as shown in the large bag (Σ6000) in Figure 6B (C), and can be regarded as a single DM signal. There is no need for , and the degree of compression is 1.

In addition, in Fig. 6B (C), the position of each bag on the horizontal axis (time axis) is as follows: bag (1) is 1/6000≦t<1/6000+1/6000x
1/6000 bag (2) is 1/6000+ 1/6000x 1/6000≦t
< 1/6000+ 1/6000x 2/6000 Bag (i) is 1/6000+1/6000x (i-1)/6000
≦t < 1/600047 + 1/6000x i/6000 bag (6000) is 1/6000+1/6000x 5999/6000≦
t < 1/6000+ 1/6000x 5999/
6000 is installed. On the other hand, the audio signals q・(t), Cl2(t) .・・・・・・、g
i(t).・・・・・・2 The time for each sampling of Q6000(t) is 1/600, respectively.
0. 1/6000+ 1/6000x 1/6000
,..., 1/6000+ 1/6000X
(i-1)/6000,..., 1/600(
)-}- 1/6000X5999/6000, each matchstick will fit into the bag one by one while touching the side of the bag. This is done for convenience; if you want to put a matchstick in the center of the bag, for example, change the time t of q1(t) to t = 1/6000 + 1/6000x (i + 0.5)
/6000 may be selected. However, even if we select 4p in this way, the conclusion of this proof will not change.

Now, compare the corresponding terms on the right sides of equations (14) and (15).

Δ・ =ΩH (1/6000+1/6000x(i
-1)/6000)g1(1/8000+1/f3
000x 3000 /6o00) (17) What the above formula means is that when sampling the audio of the first channel (CH i ), the time (1/6000+ (i-1)/6000) seconds,
(1/6000+3000/6000) seconds represents the difference in signal magnitude. This difference is like the error value of random noise, and is similar to the error value of the channel (C
H i ) i = 1, 2, ・------,
nk: may take a positive value, a negative value, or occasionally O, but generally there will be variations to the left with O as the center, and the variations will be normally distributed. Dew.

The above was related to a certain time 1=1o. The next sampling is performed after 1/6000 seconds.

The next one is delayed by an additional 1/6000 seconds. Since this sampling is performed, for example, 6000 times per second, the average value of equation (17) for one second will be (18), that is, close to O. This will become clearer if more time is spent, and it can be considered that equation (18) becomes O if the average of 10 seconds or 30 seconds is taken.

As described above, FDM and T with the same multiplicity (6000)
It has been revealed that the average power level of each CM signal is the same, and next it will be proven that the amplitude distributions of the two types of multiplexed signals are also exactly the same.

When the number of channels in an FDM signal exceeds 60, the waveform of the signal becomes almost random noise (white noise because the amplitude characteristics are flat in the target frequency band).
It is well known that this is the same as noise (also called noise).

Now, considering equation (4) above, this is not an FDM signal like equation (1), but if it is multiplexed into 60 or more call signals, each signal is a random signal.
It can be thought of as a mixed signal.

Therefore, the signal of equation (4) is expressed as the time (1/6000 + 1/6000
) , (1/6000+1/6000x9000/60
00 ) , (1/6000+1/6000x (30
H xk)/6HO) , , (k=1.2,3,
・Old...》 Signal group obtained by sampling at
Considering the amplitude distribution of Since the values were obtained by changing the sampling, it is clear that this amplitude distribution is flat regardless of time.

On the other hand, the belief @Q 1(j) shown on the right side of equation (4)
, Q2 (j) ,... CI, (t
), it will be proven below that the amplitude distribution of the signal 51 sampled at the time indicated by equation (13) is still random, and the magnitude of the amplitude is the same as equation (15).

Random signal g1 m . Q2 (j) ,...
. . , the temporal distribution of equation (14) is taken, which is a mixed signal obtained by simply changing the sampling time of each of Qo(1). That is, a group (hereinafter referred to as group B) of signal values sampled at each time given by the equation (14) below is created.

10 2 (k/6000 +1/6000x 1/6
000) Ten... Ten g6000 (k/6000 + 1/6000X 5
999/6000) (14-a) (k=2.3, 4,...) Values of many signals obtained from equations (14) and (14-a) 52 6000 ΣCli (k) (k=1.2,3,...
...) i=1? Considering the amplitude distribution, the shape of the distribution of the random signal does not change even if the sampling time is changed, and therefore the amplitude distribution is exactly the same as the amplitude distribution of the signal group in equation (4).

From the above, the FDM signal (in this case, a multiplexed signal such as equation (4) that is not expanded on the frequency axis) and the TCM signal with the same multiplicity have the same average power and the same amplitude distribution. It became clear that

The above shows that the multiple load gain obtained by FDM is ■C
This is nothing but showing that it can be obtained even with M.

However, the value of the multiple load gain cited in the above-mentioned document 3 is based on the audio signal frequency band of 0.3 to 3.4 KHz, whereas It was assumed that the same value could be obtained in the transmission band of 0.3 to 3.0 KHz. This assumption may be accepted with virtually no error.53

In addition, in the case of a TCM signal, since the signal is time compressed,
The frequency component it has becomes higher by the degree of compression, but this is because only the frequency component has been changed as mentioned above, and the signal waveform itself has simply been extended on the frequency axis, so the amount of multiload gain is Although there is no change, we will prove it strictly using mathematical formulas below.

The TCM signal is (5). It is expressed as follows using equation (6).

(19) However, eT<t<T/n+eT h・(t)=0 However, 1/n<eT<t<T+fTZ=1.2,3
,...... King is the time length of one frame. Therefore, the time-compressed CM signal is This is because only 1/n of the frame time is transmitted. If you want to express this in terms of voltage, it will be n times the power). Now, if we calculate the power in equation (20) for the time T of one frame, we get 55 (21) as in equations (7) and (8). Therefore, if the time is an integral multiple of the frame, the signal 1
{ It was found that the power possessed by (t) and G (t) is H(t)2 mi G(t)2 (22>).

Next, the frame length of the audio n-channel multiplexed TCM signal is 1.
The multiple load gain when the value is shorter than /(2fh) will be explained.

In this case, as above, FDM of audio n-channel multiplexing
It can be easily proven that it has a value equivalent to the multiload gain in the signal.

For example, assume that the frame length is 1/8000 seconds. Then, the sampling frequency was changed from the aforementioned 6000Hz to 80001-1z, and each audio signal @ql (D,
q2(1),..., CI, (t) is sampled, and the mixed audio signal G(t) is also 8000}1
2, and compare the two. That is, 56 Figures 6A (a) and (b). Figure 6B (C) (7) Just by changing the horizontal axis from 1/6000 to 1/8000,
All of the above explanations apply.

Furthermore, what happens to the multiple load gain when the frame length becomes longer than 1/(2fh) will be explained. In conclusion, the multiple load gain generally decreases, and its specific value is determined below. As a concrete value, the number of multiplexes is 6000, which is the same as the previous example, and the frame length is 1/3000.
This will be explained by taking as an example a case in which the number of seconds is reached. The size of the bags arranged on the time axis increases by the increase in one frame length, as shown in FIG. 6B (d). To be exact, the diameter of each bag is 1/6000 x 1/3000 second. Then, if we add up the diameters of all n bags, we get 1/3000
seconds. Also, the diameter of the large bag (Σ6000) is 1/30
00 seconds.

Now, as above, the sampling frequency is 17600.
Consider putting matchsticks sampled at 0 seconds into each bag. In this case, if the frame length is set to 1/6000 seconds as described above, the following inconvenience will occur. That is, FIG. 6B (d
) As shown in 1/6000 seconds, the bag is (1)
There are only up to (3000), while a matchstick has 600.
Since there are 0 bottles, each bag will contain 2 bottles. In other words, as shown in Figure 6B (d), bag (1) receives voice signals @g1 (j) and CJ2 (t), and bag (2) receives voice signals @g1 (j) and CJ2 (t).
2 m and g3 m, bag (3) has Q3 (j) and Q4 (
j)',..., hereafter, q3000 for bag (3000)
(t) and "3001 (t)" matchsticks (signal values) are put in. Note that Q1 (j) is put in the bag (6000) which is the previous sampling time.On the other hand, In bags (3001) to (6000),
No matchsticks have been sampled within this time, and the next sampling time is 17600.
Two matchsticks sampled within 0 seconds will be placed in each box.

On the other hand, the large bag (Σ6000) has two matchsticks in one frame, so two matchsticks, that is, G (1/6000+1/6000X3000/600)
0) and G (1/6000+1/6000x9000
/6000) will be inserted. However, G (1/6000+1/6000X 9000/
6000) is the matchstick (signal) sampled within the next 1/6000 second, as described above.

What does the above mean?

That is, since there are two matchsticks in the bags (1) to (3000), it means that the audio gi(j), Qi, and 1m are mixed and put into each time slot of the TCM signal. It means something that should be done. To do this technically, we need to perform 2-channel (CH) FDM and
Total of 00 slots is 2 (CH) x 3000 = 6
This means that a TCM signal of 000 (CH) should be generated. Then, the sizes of these and the one matchstick in the front of the large bag (Σ6000) are compared in the same manner as described above.

Although this is also a method, it is not a TCM signal in the original sense. Therefore, in order to include only one audio signal in each time slot, the following must be done. The 6000 audio signals are divided into two groups, one group is placed in bags (1) to (3000), and the other groups are placed in bags (3001) to (6000).
so that it can be placed in This operation will be explained using FIG. 6D (f). Two sets of bags (1) to (6000) and a large bag (Σ6000) are prepared there.

Now, the bags (1) to (6000) written on the top have matchsticks Q1, Q3. Q5. . . , g5999 are shown being addressed one by one. And the large bag (Σ6000) has G 1 (1/6000+1/
There is one matchstick indicating 6000x 3000/6000).

On the other hand, the bags (1) to (6000) written at the bottom contain one matchstick g2, g4, CJ6,...-g6000 each, and the large bag (Σ6000) contains G2.
(1/6000+ 1/6000x 9000/
There is one matchstick showing the value (6000). If you do this as shown in this diagram, you will not have two matchsticks mixed together in the same bag, and the length of one frame will be reduced to one.
By the same proof as in the case of /6000 seconds, it can be seen that the multiple load gain of the FDM signal (in this case, 3000 multiplexes) is the same as that of the TCM signal.

The above explanation can be expressed as follows.

(4) n voices (in this case n - 600
0) is divided into two and expressed as follows.

Gl (t)=g1 (t)+g3m+...+”59
99(t)(4') G2(t) =CI2(t) +Q4(t) +-+o
6ooo(t)(4'') Then, the above-mentioned proof can be performed for each of the above expressions. However, the sampling timing is assumed to be performed under exactly the same conditions as above.

In the above explanation, it has become clear that it is necessary to divide into k groups in order to obtain the multiload gain, but it will be explained below that the TCM signal requires signal compression.

In Figure 6D (f), the upper bag group (1) to (6000) and the lower bag group (1) to (13000) contain only one matchstick, but the TCM signal Considering this, there is still some dissatisfaction. That is, the upper bag group and the lower bag group are simultaneous in time, so within one time slot of the TCM signal, there are still Q, 02.1 g3, and Ω4·゜゜゜. - g5999 and g6000 are supposed to coexist. Signal compression removes this, and the following method may be used. That is, Q1
．． (;13,...' Regarding g5999, there are two bags (1) and (3001), (2) and (3002),
..., (3000) and (6000) respectively in the previous bag (1). (2) ,..., (3000), and q2, "4,"', g6000 are stored in the later bag (3001) in the same way. (3002),・
..., accommodated within (6000).

To do this, for example, for g1, it is sufficient to create a composite signal of g. That is, it is sufficient to create the sum of signals obtained at two adjacent sampling times. Technically, the audio signal is stored in a memory circuit, read out in a time-division manner at twice the speed (duty ratio 50%), and the signal obtained by sampling this read signal at a sampling rate of 1/3000 seconds is It is desired. However, in this case, the TCM signal value (voltage value) of the instantaneous value of the sampling time is
Is it impossible to faithfully reproduce the original signal with a fixed time width? time slot length).

If the above operation is viewed from the viewpoint of multiple load gain, it will be as follows.

If the frame length is longer than the sampling time interval of 1/6000 seconds, even if the multiplex number n is 6000,
The multiple load gain in FDM with 0CH multiplexing cannot be obtained, and when the frame length is 1/3000 seconds, the multiple load gain is equivalent to that of an FDM signal of 300QCH.

The multiplex load gain is determined when the frame length is further increased to 1 second and the multiplex number n = 6000.

An explanatory diagram in this case is shown in FIG. 6C (e). To explain the figure, one frame is one second, and n=600
0, therefore, 6000 bags each having a diameter of 1/6000 second are provided on the horizontal time axis, which forms one frame. In this case g1(j),...,Cl
Think about where the sampled matchsticks of 60 o (t) can be placed. Sun 63 The above frame length is 17600 for the timing of pulling.
Taking the same as the case of 0 seconds, bag (1). (2) ,
(3). ... (6000), one matchstick indicating each audio signal will be inserted into each one.

On the other hand, the large bag (Σ6000) will contain 6000 matchsticks (having values sampled at each sampling period) indicating G (t). This means that
This means that 6000 audio channels are mixed and input into each time slot of the TCM signal. The technology that makes this possible is FDM, but this does not allow for the TC
It cannot be applied to M signals. Therefore, in order to have only one audio signal in each time slot,
The 6,000 audio signals must be divided into 6,000 groups and input one group at a time (in this case, one channel at a time).

This shows that the multiple load gain in this case is O, which is not obtained at all. Further, the signal compression degree of the TCM signal in this case is 6000.

As is clear from the above explanation, we have shown that the multiple load gain decreases when the time length of one frame becomes longer than the sampling period required to faithfully reproduce the audio signal. Expressed in terms, if the frame length Tt is Tt > 1/(2 fh) and the number of multiplexes is n, the multiplex load gain is n' = nx1
/ (2fh lower.) (23) This value is equal to the multiple load gain of a frequency division multiplexed signal having a multiplex number determined by the following value.

According to the above-mentioned documents 1 and 2, the practical range of the frame length T1 and the number of multiplexes n is as follows: frame length T: 0.1 seconds≧Tt≧0. 00
1 second t Multiplex number n: about 3000≧n≧2, and it is clear from the above example that as long as it is within the above range, equation (23>) always holds true.

In addition, in order to address one matchstick in each bag as a TCM signal, in other words, to address one original audio signal, if one frame time length Tt is longer than 1/2ffi, the signal It has also become clear that the compression ratio ψ is given by ψ=Tt/(1/2h).

Note that although the above description assumes that the multiplicity is 60 or more, a case where the multiplicity is less than 60 will be described. At this time, the number of bags within one frame of the TCM signal is 60 or less.

On the other hand, since the FDM signal has a multiplicity of 60 or less, the signal cannot be considered as random noise.

That is, although the FDM signal or the signal shown on the right side of equation (4) may be said to be a random signal approximately, a 1.1 error will occur in the result obtained as a random signal. However, in reality, whether or not it is a random signal has no relation to the multiple load gain, and we have already explained the characteristics of two signal value groups A and B obtained when the sampling time of the multiplex signal is shifted. The method can now be applied again for multiplicity 60 or less to prove that groups A and B have the same signal magnitude and distribution.

First, the fact that the average magnitude (power) of the average amplitude of group A is equal to that of group B can be achieved by similar derivation from the sampling theorem already explained. Next, regarding the amplitude distribution, since audio signals do not always have a certain periodicity, there is no reason why there should be a difference between the amplitude distribution of group A and that of group B.

As described above, when the multiplicity is 60 or less, the multiload gain of the CM signal is equal to the multiload gain of the FDM signal with the same multiplicity. However, when determining multiple load gain, TC
For the multiplicity of the M signal, the value given by equation (23) must be used.

(4) Regarding the multiple load gain of TCM signals received at the radio base station 30 The radio base station 30 receives TCM signals transmitted from a large number of mobile radios 100, but the received waves have Consider multiload gain. To conclude, as will be described later, the mobile radio device 100 transmits the information to the radio base station 3.
Assuming that exactly the same multiple load gain as when transmitting from 0 can be obtained, it is understood that transmission may be performed with a deeper modulation factor.

a 7 As a specific example, one frame length is defined as sampling time interval 1
/6000 seconds and the number of multiplexes is 6000. It is assumed that the wireless base station 30 is communicating with 6000 mobile wireless devices 100 simultaneously using the same carrier wave and using all 6000 time slots of one frame.

The positions of the mobile radio devices 100 are arranged at equal intervals on the same circumference when viewed from the radio base station 30, and the receiving antenna of the radio base station 30 is omnidirectional, and the transmitting antenna of the mobile radio device 100 is also omnidirectional. And the magnitude of the transmission power from each mobile radio device 100 is all the same, and each mobile radio device 100
It is assumed that the carrier waves used for transmission are phase-synchronized with each other. Furthermore, the radio wave propagation characteristics between the mobile radio device 100 and the radio base station 30 are
and the wireless base station 30 are identical. Under the above assumptions, the transmitted signals from each mobile radio device 100 that enter the radio base station 30 will be received exactly the same. Therefore, the state of the received signal within one frame in this case can be considered to be exactly the same as when it is transmitted from the wireless base 830. Conversely, even though each mobile radio device 100 transmits only a single voice channel in the time slot given to it, the number of multiplexed channels is 600, assuming that a multiple load gain can be obtained.
This indicates that transmission may be performed using a modulation depth that allows for a multiload gain of 0.

Although ideal conditions have been set above, consider the actual system operating conditions. In this case, the positions of the mobile radio devices 100 are randomly scattered, and the radio wave propagation state changes in various ways, so the received power of the radio base station 30 varies for each time slot. Furthermore, the carrier waves from each mobile radio 11100 are not necessarily phase-synchronized. Therefore, if the depth of modulation is large in a time slot where the received level is high, it is expected that the adjacent time slots will be adversely affected due to the influence of multiplex radio wave propagation. However, this can be alleviated by taking other measures such as increasing the guard time.

In addition, in the case of the small zone system, there is no need to worry too much about the possibility that the transmitted wave of one mobile radio device 100 will cause interference to another radio base 830 that is located at a different location, as co-channel interference is 0Φ. There is a possibility that this method can be used to gradually reduce the number of repeated zones. It's FM (P
M) This is because multiple load gain is used as modulation and as a result of deep modulation, wideband gain can be obtained and resistance to co-channel interference is increased.

Taking all of the above into consideration, the system design assumes that substantially the same multiple load gain can be obtained when the mobile radio device 100 transmits and when the radio base station 30 receives, as when transmitting from the radio base station 30. It became clear that it could be done.

(5) Specific method of utilizing the multiple load gain The multiple load gain in the case where one frame length is 1 msec and the number of multiplexes of TCM (time division time compression multiplexing) is 500 is determined, and an example of its use will be explained.

First, S C P C (Sinole Channel
el Per Carrier, 1 phone per carrier
The signal S to noise N ratio of analog FM (in which the channel signal is modulated) is determined, and this is compared with TCM. The level (voltage value) of the input signal to the receiver is C, the FM modulation index is mf, the noise level (voltage value) per unit frequency is n, the bandwidth of the low frequency amplifier in the discriminator output stage is Fa,
If the highest modulation frequency f, of the modulated wave is equal to [,
The signal-to-noise ratio is given by the formula below.

S/N=3"2m G(2F)-"'f
a (24) This formula was quoted from the following literature.

Edited by Sugawara “FM Radio Engineering” Published by Nikkan Kogyo Shimbun, 1952
Year 401 page (13.25) Formula Next, TC of multiplex number Q
The S/N when the M signal is FMed under the following conditions is:
It is given by the following formula.

(25) However, 71 Fao10F, mH)=Qm((2 5')n
Q = Q n In other words, in the TCM signal, the frequency of the original signal is multiplied by Q, so the bandwidth of the low frequency amplifier increases by Q times, and the depth of modulation (modulation index) also increases by Q times, so the noise Since the level band is also multiplied by Q, it is appropriate to set it as shown in equation (25).

Next, the TCM received signal level (26) for the S/N on the left side of equations (23) and (24) to take the same value. From equation (26), C, =Q" C (27) This means that the reception level needs to be Q1/2 times higher in terms of voltage and Q times as much as the power level.Therefore, the transmission power needs to be increased by Q times higher than the S-end CPC.

Next, find the multiple load gain of the TCM signal in the above example.

As already explained in (3), the FDM equivalent multiplex number in this case is n' = 500xl/6000 (m sec> -
'． - (1/1000 (msec)) 500xl/6 = 83 CH Therefore, from FIG. 8, the multiple load gain is 28. It can be seen that the value is between 6 dB and 32.6 dB for 120 channel multiplexing.

If the graph shown in FIG. 7 is created and estimated based on FIG. 8, a multiple load gain of 30 dB@- can be obtained. Therefore, this multiple load gain is used to deepen the modulation depth (deviation) and reduce the transmission power. Assuming that the transmission output of a non-TCM SCPC, that is, a 1-channel analog FM signal, is 10 mW at the cordless telephone level, the required transmission power in this case can be found by multiplying by 500 using formula (27) and then subtracting the multiple load gain. , (10mWx 500)-30dB=10mWx
We get 500/1000=5.0mW (28). In other words, using TCM requires less power.

Next, the physical meaning of multiple load gain in TCM signals will be explained, and points to be noted when using this as a system will be described.

The TCM signal has a long frame period of 1 second or more (multiplexing number 6).
000), no multiload gain can be obtained; however, in this case, the FM of the TCM signal
The modulation index has a constant value defined by the system. For example, the modulation index of the original signal @ (0.3 to 3.0 KHz') is 1.75 KHz (in the case of an IKHZ tone signal with standard modulation deviation), and in the case of TCM where 500 of these are multiplexed, the signal band is 150 ~ 1500KHZ
, the standard modulation deviation is 875K}-12 ( 500K
(When HZ tone signal is standard modulated). However, if the frame length is set to 1 msec, a multiple load gain of 30 dB can be obtained as described above, and this multiple load gain is used to increase the depth of modulation, but what will happen to the actual modulated wave? I'll explain if there is one.

First, consider an all-channel implementation, ie, a case where telephone signals are flowing in all C time slots. In this case, the effect of a multiple load gain of 30 dB on the modulation shift increase is:
Considering from Fig. 9, the multiple load gain is the value obtained by subtracting about +8 dB from the relative power of the sine waves with the same peak value, so one time slot can be obtained by using a CM signal with an arbitrary frame length. It can be seen that it is equal to the modulation shift of the signal when using only

Next, consider the case where the load is gradually reduced from all time slot implementation. That is, time slot 1~
The problem is what will happen to the modulation shift of the signal if some of the time slots are used for actual audio signal transmission and the rest are not used as empty time slots.

In this case, since the number of implemented channels is reduced, the multiple load gain is naturally reduced as well. For example, in the case of 250 implementation of 75 1/2, the multiload gain is n' = 250xl/6000-:- (1/100
0)-250X 1/6 =/12 (G l-1
> Therefore, from Fig. 7, the multiple load gain is 24.5d.
It turns out that it is B. However, since the load is 172, the power level of the modulated signal is lowered by 3 dB. Therefore, the equalizing multiple load gain in this case is 27.
Although the effective modulation depth of CM-FM is 5 dB, it is considered to have no effect on system operation.

Furthermore, the effective modulation shift is determined when the number of implementations is reduced and only one time slot 1~ is used. Although the multiple load gain R of one channel is OdB, the signal load is reduced by 1/500, that is, 27 dB, compared to when all the components are installed. Therefore, the apparent multiple load gain is 27dB.
Therefore, even if the modulator is operated with this value set to 30 dB, it is assumed that there is no effect on the system operation. -C is good. In addition, in the manual stage of the modulation circuit of an actual radio, the IDC (
Instantaneous Dev+at+on C
ont7th rol WRFfJ modulation deviation amount suppression circuit is provided with a function to limit the depth of modulation to a certain value or less. Therefore, the modulator output is DCM
Regardless of the implementation status of the telephone signal g, the effective modulation deviation must be kept below a certain value.As explained above, the multiple load gain of the TCM signal
It has become clear that by using this method to increase the modulation deviation of the M signal, it is possible to significantly reduce the transmission output. Technically, this means that it has a very large effect on power saving. Of course, with SCPC, a wireless device with continuous transmission of 10 m is operated at a time rate of 1/500, that is, 0.2%, and its output is 10 mW, which is 172 times 5 mW, so it has a power saving effect. It is obvious that .

The advantages and disadvantages of an example in which a guard time is provided between time slots of a TCM signal according to the present invention will be explained.

The TCM signal described in this section does not necessarily require a guard time between pulse trains like a digital signal. However, a guard time may be provided between time slots in order to eliminate the influence of timing deviations of synchronization signals and delayed waves due to multiplexed waves on radio wave propagation. Although the specific value of the guard time varies depending on the system to which it is applied, for example, 0.1.1-0.5 μsec is appropriate for an indoor mobile phone system, and 5 to 10 μsec is appropriate for a car phone.

In a system where a guard time is set, if the frame length is constant, the slot time is increased by the amount of the card time.
Because the Ffl width of the time decreases, the compression ratio of the original signal must be increased; therefore, the highest frequency of the signal becomes higher. In the cordless telephone example above, the time slot is lmsec : 500=2 μsec
・10%, that is, 0.2μsec guard・
Taking the time, the time slot is 1.8 μsec. Also, the highest frequency is 3 without guard time.
KllZ x 500 = 1.5M) - 1z xlO/9 if you take 10% guard time from 1.5MHz
= 1.67 MHz. Therefore, the required bandwidth becomes correspondingly wider, and the frequency effective utilization rate decreases by 11%.

Next, apply multiple load gains to the amplifier design. In this case, the level of the multiplexed voice converted into TCM may be considered to be lower than the level conventionally considered by the amount of the multiplex load gain. Therefore, even if the amplification factor can be increased by that amount, or the output level can be made higher than the conventional one by the amount of the multiple load gain, the distortion factor, etc. will remain at the value conventionally assumed.

Multiple negative gain can be applied not only to active circuits as described above, but also to
It is also applicable to passive circuits as described below. In other words, if applied to a mixer circuit, even if the rated output is increased by the amount of the multiple load gain, it is possible to operate in the conventionally assumed operating state. When applied to a wireless transmitter, this has the following benefits. For example, the first
Inserting a power amplifier into the output of the transmission mixer 133 in Figure B is often used to increase the reach of radio waves. In this case, by introducing the multiple load gain, it is possible to increase the 7q transmission output level higher than the level conventionally assumed by the amount indicated by the multiple load gain.

Alternatively, if the same transmission level as the conventional one is sufficient, the rated output of the amplifier can be made with a lower level output than the conventional one by the amount of the multiple load gain.

The above concept of rated power can be applied not only to transmission mixers but also to all resistors, capacitors, inductances, etc.

[Effects of the Invention] As is clear from the above explanation, the multiple load gain of time-division time compression multiplexed signals, which had not been clearly shown in the past, was categorically clarified using system parameters, and as a result, for example, Even if the angle modulation depth (deviation) is increased by the amount of multiple load gain and transmitted, the influence on the wireless channel can be suppressed within the conventional design value, and one wireless channel This makes it possible to reduce the per-transmission output level compared to conventional systems, resulting in power savings, as well as rational and economical methods for designing amplifiers and determining the ratings of passive elements. Since the design has become possible, the effect on communication systems, especially wireless systems, is extremely large.

[Brief explanation of drawings]

Figure 1A is a conceptual block diagram showing the concept of the system of the present invention, Figure 1B is a circuit diagram of a mobile radio used in the system of the present invention, and Figure 1C is a wireless base used in the system of the present invention. FIG. 2A is a time slot structure diagram for explaining the time slots used in the system of the present invention; FIG. 2B is a diagram showing the radio signal waveform of the time slot; 3B and 3B are spectrum diagrams showing the spectra of speech signals and control signals, FIG. 3C is a circuit configuration diagram for multiplexing voice signals and data signals, and FIGS. 4A and 4B are diagrams showing the operation of the system according to the present invention. A flow chart showing the flow, FIG. 5 is a spectrum diagram of a frequency division multiplexed signal, FIG. 6A is an amplitude diagram showing changes in amplitude of a time division time compression multiplexed signal, and FIGS. 6B, 6C, and 6D are A sampling diagram showing how the time division time compression multiplexed signal is sampled; Figure 7 is a diagram showing the relationship between the multiple load gain of the time division 05 compression multiplexed signal and the number of multiplexed audio signals; Figures 8 and 9; is a multiple load gain diagram illustrating the relationship between the multiple load gain of a frequency division multiplexed signal and the number of communication paths, cited from a known document. O...Telephone network 20...Kanmon switchboard 2-1
~22-n...Communication signal O...Radio base station 1...Signal processing section 2...Wireless transmitting circuit 35...Radio receiving circuit 8.
...Signal speed restoration circuit group 38-1 to 38-n...Transmission speed restoration circuit 39...
Signal selection circuit group 39-1 to 39-n...Signal selection circuit 40...Control section 41...Clock generator 42...Timing generation circuit 51...Signal speed conversion circuit group 51-1... 51-n...Signal speed conversion circuit 52...
Signal assignment circuit group 52-1 to 52-n...Signal assignment circuit 91...Digital encoding circuit 92...Multiple conversion circuit 100, 100-1 to 100-n...Mobile radio device 10
1... Telephone unit 120... Reference crystal oscillator 121-1, 121-2... Synthesizer 122-1
, 122-2...Switch 123...Transmission/reception intermittent controller 131...Fast/reverse conversion circuit 132...Wireless transmission circuit 133...Transmission mixer 8
3 134...Transmitter 135...Radio receiving circuit 136...Receiver Miku'J- 137...Receiver 138...Speed recovery circuit 141...Clock regenerator.

Claims (1)

[Claims] 1. Each radio base means (30) each covering a plurality of zones to constitute a service area, and a radio base means (30) for moving across the plurality of zones and communicating with the radio base means; barrier exchange means (20) for exchanging communications with each mobile radio means (100) using a radio channel carrying time-compressed delimited signals in time slots of a frame structure; In mobile communication methods,
Time-division communication for mobile communication, wherein the level of a radio signal used for communication between the radio base means and the mobile radio means is determined based on the multiload gain obtained by the time-compressed segmented signals. Method. 2. Each radio base means (30) each covering a plurality of zones to constitute a service area, and a time slot arranged in a frame for moving across said plurality of zones and communicating with said radio base means. In a mobile communication system using a barrier exchange means (20) for exchanging communication between each mobile radio means (100) using a radio channel carrying time-compressed segmented signals, Time division of mobile communication, wherein at least one of the radio base means and the mobile radio means has its transmission power level determined based on the multiload gain obtained by the temporally compressed segmented signals. Communications system.