KR20190078832A - MAC protocol for wireless sensor network - Google Patents

Abstract

The present invention relates to a QCA decoder. The QCA decoder according to the present invention selects a Montgomery factor and derives a multiplication formula on a finite field GF (2 ^m ) using the selected Montgomery factor. Then, the derived multiplication expression is separated into two expressions in which there is no data dependency, and operations corresponding to the two separated expressions are performed in parallel. According to the present invention, a new multiplication method for Montgomery multiplication in a polynomial basis on GF (2 ^m ) is provided, and a semi-systolic multiplier capable of parallel operation based on this multiplication method can be provided.

Description

MAC protocol for wireless sensor network {MAC protocol for wireless sensor network}

The present invention relates to a wireless communication method, and more particularly, to a wireless communication method using an asynchronous MAC protocol for a wireless sensor network.

The present invention is derived from the results of the research project of the SW-centered university of the Ministry of Science, Technology, Information & Communication and the Information and Communication Technology Promotion Center (Project No. 2017-0-00096)

Unlike other networks such as WLAN and Bluetooth, sensor nodes in the Wireless Sensor Network (WSN) can not change or charge their own batteries. Thus, energy efficiency is a fundamental and important consideration when designing protocols for sensor nodes in wireless sensor networks. Energy constraints, along with the large number of nodes, make designing routing protocols for sensor nodes, MACs, and wireless sensor networks difficult.

Many researches have been conducted to reduce the energy consumption of wireless sensor networks. The most common mechanism for reducing energy consumption in wireless sensor networks is duty cycle. Duty cycling techniques conserve energy by switching sensor nodes between awake and sleeping states. Energy-efficient conventional duty-cycling MAC protocols can be divided into two types: synchronous and asynchronous protocols.

Wei Ye, John Heidemann, and Deborah Estrin, "An Energy-efficient MAC Protocol for Wireless Sensor Networks," 21st International Annual Conference of the IEEE Computer and Communications Societies, pp. 1567-1576 , June 2002), Tijs van Dam and Koen Langendoen, " An adaptive energy-efficient MAC protocol for wireless sensor networks ", Proc. Of the first ACM Conference on Embedded Networked Sensor System, pp. 171-180, Nov. 2003), the active period of each node's duty cycle is synchronized by exchanging periodic control frames such as SYNC. These additional control frames cause wasted energy usage. Further, when common periods for accessing the medium overlap, collision between frames occurs.

In contrast, the B-MAC (Joseph Polastre, Jason Hill, and David Culler, "Versatile low power media access for wireless sensor networks," Proc. Of the ACM Conference on Embedded Networked Sensor Systems, pp. 95 -107, 2004) and Michael Buettner et al., "X-MAC: a short preamble mac protocol for duty-cycled wireless sensor networks," Proc. of the 4th ACM Conference on Embedded Networked Sensor System, pp. 307-320, Nov 2006) are more energy efficient than synchronous protocols.

This protocol uses preamble sampling instead of an additional sync control frame. As the number of control frames is small and the possibility of collision is low, the asynchronous MAC protocol is more effective in terms of energy consumption. Even though the asynchronous MAC protocol is carefully designed to operate without energy wastage, it is quite difficult to match the awake state between the sender node and the receiving node.

Therefore, it is necessary to consider the asynchronous sensor MAC protocol, which is further improved in terms of delay, throughput, and energy consumption, and a wireless communication method using the same.

It is therefore an object of the present invention to provide an asynchronous MAC protocol for an improved wireless sensor network in terms of delay, throughput and energy consumption, and a wireless communication method using the asynchronous MAC protocol.

According to an aspect of the present invention, there is provided a QCA decoder including a first wiring for receiving a first input signal, a second wiring for receiving a second input signal, a third wiring for receiving an enable signal, Fourth and seventh wires connected to the second wire to transmit a signal, a fifth wire connected to the first wire to transmit the first input signal, and a second wire connected to the second wire to transmit the signal, A seventh wiring connected to the first wiring, a fourth wiring connected to the fifth wiring, and a third wiring connected to the third wiring so that a signal obtained by inverting the first input signal is transmitted, A first combinational logic cell that performs a logic AND gate function to output a first output signal, a fifth combinational logic cell connected to the fifth wiring, the sixth wiring, and the third wiring, A second combinational logic cell for outputting a signal, A third combinational logic cell connected to the seventh wiring, the eighth wiring, and the third wiring for performing a logical AND gate function to output a third output signal, And a fourth combinational logic cell coupled to the third interconnect to perform a logical AND gate function to output a fourth output signal.

The first through fourth combinational logic cells may be a 5-input majority-gate structure having two fixed inputs.

The first to third wirings are made of rotated cells, and the second wirings are arranged substantially parallel to the first wirings, and the third wirings are substantially parallel to the first and second wirings And the fourth to ninth wirings are made of general cells and can be arranged at substantially right angles to any one of the first to third wirings.

In addition, any one of the first to fourth output signals may be generated after four clock phases.

According to another aspect of the present invention, there is provided a quantum dot cellular automata device including the QCA decoder.

According to the present invention, it is possible to provide a RIX-MAC protocol superior to the existing X-MAC protocol in terms of delay, throughput and energy consumption by avoiding collision and reducing overhead. In addition, using the IX-MAC protocol according to the present invention, it is possible to perform rapid communication in the wireless sensor network without sacrificing energy saving in both the transmitting node and the receiving node, and to reduce the collision and unnecessary wakeup, .

1 is a drawing referred to in the description of operations for X-MAC,
FIG. 2 is a block diagram illustrating the operation of the RIX-MAC protocol according to the present invention.
3 is a timing diagram of the RIX-MAC protocol according to the present invention,
4 is a diagram referred to in the description of random backoff in the RIX-MAC protocol according to the present invention,
5 is a diagram illustrating a Markov chain model of a node in the RIX-MAC protocol according to the present invention, and
FIGS. 6 to 14 are diagrams showing the results of simulation for the RIX-MAC protocol and the X-MAC protocol according to the present invention.

Hereinafter, the present invention will be described in detail with reference to the drawings.

1 is a drawing referred to in the description of operations for X-MAC.

The X-MAC proposed by Michael Buettner et al. Is an asynchronous duty-cycle MAC protocol for B-MAC based wireless sensor networks proposed by Joseph Polastre et al., Which uses a series of short preambles to implement low power communication without synchronization. Since the short preamble carries the address of the receiving node, other nodes than the receiving node can sleep while listening to the first short preamble. The receiving node may respond with an early-ACK to stop the preamble and start data transmission. The main features of X-MAC are short preamble sampling and early-arc mechanism. The X-MAC transmits multiple short preambles so that the receiver notices instead of the long preamble used in the B-MAC. If the receiver detects a short preamble during an awake period, it sends an early acknowledge to the transmitter before receiving the actual data frame, and the transmitter knows that the receiver is staying in an awake period, .

As a result, X-MAC saves more energy than B-MAC due to short preamble sampling and early-arc mechanism.

As shown in FIG. 1, all nodes in the network have their own wakeup schedule to transmit and receive frames with a period T. The TX node continues to send a short preamble until the RX node wakes up. The other node does not interfere with the communication between the TX node and the RX node. When the TX node receives the acknowledgment, it begins to transmit data frames to the RX node.

Since the X-MAC is asynchronous, the average communication time of the TX node is T / 2 + data frame length. When more than one TX node is woken up and simultaneously starts to send a preamble, all other nodes, including the RX node, can not verify the address information of the preamble, so the TX node does not interrupt the preamble transmission until the next wakeup period. Thus, for each collision data frame, the average communication time for the sender is T.

FIG. 2 is a diagram referred to in describing the operation of the RIX-MAC protocol according to the present invention.

The RIX-MAC protocol according to the present invention is based on X-MAC. Every node in the network has the same cycle as the T cycle divided into two states: wakeup and sleep state. The wakeup state is an active state in which a node turns on wireless communication and receives or transmits data, and a sleep state is a state in which a node turns off wireless communication to conserve power. The wakeup state is divided into two periods: a synchronization wakeup period when the node tries to transmit a data frame, and an optional scheduling wakeup (scheduled wakeup) and a synchronization wakeup (synchronized wakeup) period.

Figure 2 shows the basic operation of the RIX-MAC protocol, where the TX node transmits a short preamble and wakes up twice. Each node wakes up periodically according to its schedule and checks whether there is a short preamble coming in. If there is no short preamble after scheduling, the TX node goes into a power saving state. When there is a data frame to transmit, the TX node wakes up again in Synch-wakeup synchronized to the RX node. When the channel is idle, the TX-node starts transmission of the short preamble before transmitting the data frame.

In the RIX-MAC protocol, all nodes independently have their own wakeup schedules, have a temporary wakeup schedule synchronized to the wakeup of the RX node, and the wakeup schedules of the RX node are obtained as wakeup times within the fields of the early- . Therefore, a TX node using RIX-MAC wakes up periodically twice when there is data to be transmitted from TX node to RX node. At the TX node, the first wakeup receives a data frame according to its own schedule (Sched-wakeup), and the second wakeup transmits data to the RX node (Synch-wakeup) as shown in Fig.

Since the TX node initially does not have information about the wake up schedule of the RX node, the RIX-MAC acts like an X-MAC during the first cycle of transmission. When the TX node has data to send to the RX node, the TX node retrieves the wakeup schedule of the designated RX node in the table. If an appropriate wakeup schedule for the RX node is not found, the TX node continues to transmit a short preamplifier in its own schedule as in X-MAC. Upon receipt of an early acknowledgment from the RX node, the TX node immediately begins transmitting the data frame, extracts the wakeup information of the RX node in the wakeup time field of the early-ag frame, and transmits the wakeup information to the up-wakeup and Synch-wakeup In a wake-up time table as a time difference between them.

3 is a timing diagram of the RIX-MAC protocol according to the present invention.

사이의 간격으로 결정된다. 데이터 도착 시간 D가 상위 계층으로부터 MAC 계층에 도달하는 순간을 의미하면, 시간 간격

은

와

로 표시된 데이터 도착시간 D에 의해 두 개의 주기로 나뉜다. 마지막 Sched-wakeup에서

내에 MAC 계층에서 데이터가 준비되지 않으면 다음주기에 Synch-wakeup이 시작됩니다.Referring to FIG. 3, when the TX-node predicts the wakeup of the RX node, the Synch-wakeup detects the time interval between the last sched-wakeup and the RX node

As shown in FIG. If the data arrival time D is the moment when the data arrives from the upper layer to the MAC layer,

silver

Wow

The data arrival time D shown in FIG. From the last Sched-wakeup

If data is not ready in the MAC layer within the next cycle, Synch-wakeup will start.

When multiple TX nodes attempt to transmit data frames in the next Sched-wakeup of the RX node to adjust the preamble transmission, an additional field, a duration, is inserted in the preamble and early-arc frame. This field contains information about the number of slots needed to transmit a data frame for each TX node, and other nodes can predict the duration of the next data transmission. The RX node uses the duration field in the early-arc frame to coordinate the transmission of the short preamble contending for multiple TX nodes.

A TX node that reads the duration value at another TX node or RX node sets a network allocation vector (NAV) to delay media access and stops to reduce the number of slots. After successful transmission of the data frame, the TX node under the virtual carrier sense with NAV resumes its decreasing slot count and attempts to transmit a short preamble.

Since the TX node tries to transmit the wake-up of the RX node, collision problems can occur due to the receiver-initiated protocol when multiple nodes attempt to transmit data to the RX node. To avoid preamble collisions by multiple senders, the TX node initiates a random backoff procedure after the Synch-wakeup. The random backoff assigned to the TX- node means the number of time slots waiting for the next transmission of the short preamble after Synch-wakeup.

4 is a diagram referred to in the description of random backoff in the RIX-MAC protocol according to the present invention,

In FIG. 4, TX1 and TX2 indicate that the transmitter R and TX1, which attempt to transmit data to the receiver, precede TX2 to occupy the medium. TX1 captures the medium to transmit the short preamble and RX response to the early-ack preamble. TX2 receives the early acknowledgment from RX and sets NAV to the end of the data transmission. The duration of the early ACK frame specifies the number of time slots that TX2 should wait during data transmission. When the data transmission is completed, TX2 again counts the remaining slot count and transmits the preamble in the next time slot.

In order to model a sensor node having a finite queue in a duty cycle sensor network, a Markov chain having a finite state number indicating a queue length at the time of node wakeup can be used. In this model, it is assumed that the frame arrives independently for each node, the channel is ideal, retransmission is not allowed, and each node can buffer the frame in a finite queue. The present invention also assumes that there is only one receive and receive per node during the duty cycle and the transmission probabilities of all nodes are constant. This model is used to calculate the throughput, delay and energy consumption of the RIX-MAC.

5 shows a Markov chain model of a node in the RIX-MAC protocol according to the present invention.

FIG. 5 shows a Markov model with a queue capacity Q, where each state represents the number of data frames in the queue. If the queue is not empty, the node attempts to access the media and transmit the data frame. When a queue is transmitted, the data frame is removed from the queue, and if the queue overflows, the data frame is deleted. If A_i is the arrival probability of i data frames at the node and p is the probability of transmitting frames in the cycle, the transition probability from one state to another is:

이다.here,

to be.

이다.here,

to be.

가 상태 공간에서 상태 i의 정적 확률인 경우,

, 마르코프 모델은 환원 불가능하고 비 주기성이므로 고정 분포

가 존재한다.Specifically, [Equation 2] and [Equation 5] show that the frame is deleted when the queue overflows.

Is a static probability of state i in state space,

, The Markov model is irreducible and non-periodic,

Lt; / RTI >

에 대하여,

about,

And

가 천이 행렬이다.here,

Is a transition matrix.

와

는 p의 함수로 나타낼 수 있다. 따라서 각 노드에 대한

와 p의 관계는 함수

,If frame arrival information is known, the probability p for each node to acquire contention is the only variable in the transition matrix

Wow

Can be expressed as a function of p. Therefore,

And p is the function

,

은 제안된 마코브 모델을 사용하여 각 노드가 주어진 확률 p에 대해 경쟁을 획득 할 수 있음을 알 수 있다. [수학식 8] 이외에,

와 p 사이의 다른 관계는 MAC 프로토콜의 미디어 액세스 규칙에서 얻을 수 있는

과 p 모두를 해결하는데 필요하다.Fixed probability of empty queue state

We can see that each node can obtain a competition for a given probability p using the proposed Markov model. In addition to the equation (8)

And p can be obtained from the media access rules of the MAC protocol

And p.

는 사이클의 활성 기간, τ는 타임 슬롯이다. RIX-MAC에서, 모든 노드는 자신의 스케줄 (Sched-wakeup)에서 주기적으로 깨어나고 깨어 일어나 수신기의 웨이크 업 (Synch-wakeup)에 동기화된 데이터 프레임을 전송한다. RIX-MAC은 둘 이상의 송신기가 깨어나고 동일한 백 오프 후에 자신의 프리앰블 전송을 시작할 때 충돌을 일으킬 수 있다.The RIX-MAC protocol according to the present invention is an asynchronous time slot protocol, T is the length of the cycle,

Is the active period of the cycle, and [tau] is the time slot. In RIX-MAC, all nodes wake up periodically in their sched-wakeup and transmit synchronized data frames to the receiver's wake-up (Synch-wakeup). RIX-MAC can cause collisions when two or more transmitters wake up and begin their own preamble transmission after the same backoff.

수신자에 대한 프레임은 다음에 의해

함수로 기술될 수 있다.Assuming that there are N nodes in a fully connected network, when a node has a frame of data to be transmitted, and when a j node has a frame to transmit in a queue competing to transmit the kth node, k nodes out of j nodes (Synch-wakeup) probability

The frame for the recipient is

Function.

,

,

는 다음과 같이 계산될 수 있다.k nodes compete for a channel by waking up in sync with the receiver and sending a preamble frame. The probability of successfully delivering the preamble frame to the receiver

Can be calculated as follows.

이다.here,

to be.

는 다음과 같이 도출될 수 있다Where W is the size of the competing window. Therefore, the probability that each node gets a channel p and the probability that each node will successfully transmit a data frame

Can be derived as follows

을 구할 수 있고 [수학식 12]에

을 연결하면 각 노드가 데이터 프레임을 성공적으로 전송할 확률,

를 결정할 수 있다.

, p 및

를 이용하면 네트워크의 평균 처리량과 지연을 얻을 수 있다.Using Equations (11) and (7), the fixed probability

≪ EMI ID = 12.0 >

The probability that each node successfully transmits a data frame,

Can be determined.

, p and

The average throughput and latency of the network can be obtained.

Throughput is defined as the amount of data successfully delivered within the elapsed time. Since the RIX-MAC protocol according to the present invention operates with a duty cycle, the throughput can be calculated within the cycle time. Therefore, the average throughput of RIX-MAC is

Where S is the size of the frame.

와 경쟁 지연

와 같은 두 가지 구성 요소로 구성되어

가 된다. 대기열 지연은 모든 이전 프레임이 전송될 때까지 프레임이 대기열에 있어야 하는 시간이다. 경합 지연은 [수학식 10]에서와 같이 대기열로 들어오는 프레임이 대기열에서 전송된 프레임의 시간까지의 간격으로 정의됩니다. The delay experienced by the frame is the wait delay

And delayed competition

It consists of two components:

. The queue delay is the time frame must remain in queue until all previous frames have been transmitted. The contention delay is defined as the interval between the frames arriving at the queue and the time at the frame transmitted from the queue, as shown in [Equation 10].

During the T period,

는 선행 데이터 프레임의 경합 지연

의 함수에 의해 획득될 수 있다. 데이터 프레임은 큐의 헤드에 있는 프레임이 전송 매체에 대한 경합 프로세스의 중간에 있을 때 큐에 들어갈 수 있으며 경합 지연의 절반 동안 평균적으로 대기해야 한다. 경합 지연

를 얻기 위해, 데이터 프레임을 갖는 노드는 T의 사이클에서 미디어에 대해 한 번 경합하며, 여기서 노드는 경합을 획득하기 위해 p의 확률을 갖는다. 따라서 데이터 프레임에 대한 전체 지연은 [수학식 14]와 [수학식 15]의 두 지연 요소의 합으로 계산된다.By the Markov model, the queuing delay for the data frame

The contention delay of the preceding data frame

Lt; / RTI > The data frame may enter the queue when the frame at the head of the queue is in the middle of the contention process for the transmission medium and it should wait on average for half of the contention delay. Contention delay

, A node with a data frame competes once for the media in a cycle of T, where the node has a probability of p to obtain contention. Thus, the total delay for the data frame is computed as the sum of the two delay elements of (14) and (15).

Since the RIX-MAC protocol according to the present invention is a duty cycle protocol, the average energy consumption P per cycle can be obtained from a value obtained by dividing the energy consumption E of a node in a cycle by a cycle length T. [ The RIX-MAC protocol has two active sections, Sched-wakeup and Synch-wakeup, and the Synch-wakeup section is optional when the node wants to transmit data frames. If only one node in the network has a data frame addressed to the RX node during a periodic wakeup or if two or more nodes have different backoff values and a data frame destined for the RX node, Can successfully receive a data frame. If the RX node has a data frame with more than one node having the same backoff value and pointing towards the RX node, the RX node will not receive the data frame.

Thus, in Synch-wakeup, the TX node performs one of the following actions:

의 확률로 데이터 프레임의 성공적인 전송One)

Successful transmission of data frames with probability of

의 확률로 데이터 프레임의 충돌 송신 는2)

The collision transmission of the data frame with the probability of

확률로 전송하지 않고 슬립3)

Slip without probability transmission

 The Sched-wakeup RX node behaves as one of the following:

의 확률을 갖는 데이터 프레임의 성공적인 수신One)

Successful reception of a data frame with a probability of < RTI ID = 0.0 >

확률로 데이터 프렘임의 충돌2)

Chance to crash data prime

의 확률로 액티브 상태 머물면서 유휴 상태3)

Probability of staying active and idle

,

및

로 표시한다. 우리는 노드가 프레임을 전송하고, 프레임을 수신하고,

,

및

로 표시되는 휴면 (sleep) 전력을 갖는다.To calculate the energy consumption, we set the duration of the preamble, data frame and early arc frame to

,

And

. We assume that a node sends a frame, receives a frame,

,

And

Quot; sleep "

의 시간이 소요되며, 여기서

는 랜덤 백 오프의 평균 지속 시간이다. 노드가 Synch-wakeup 에서 데이터 프레임을 성공적으로 전송하는 에너지 소비량

은 다음과 같다.Thus, for each case, we can calculate the node and obtain the average energy consumption of the node during the period. First, we will derive the energy consumption of the TX node during the sync-wakeup period. On average, to transmit data frames

Of time, where

Is the average duration of the random backoff. Energy consumption by which a node successfully transmits data frames in Synch-wakeup

Is as follows.

은 다음과 같다.When a preamble collides in a Synch-wakeup, it can not receive a preamble frame, and the TX node continues to transmit the preamble during the remaining active period. Thus, the energy consumption that the node transmits data frames and encounters collisions in Synch-wakeup

Is as follows.

The energy consumption of the sleep node is assumed to be negligible so that E_S is zero.

이다. 노드가 Sched-wakeup에서 데이터 프레임을 성공적으로 수신하면 에너지 소비량

는 다음과 같다.During the scheduled wakeup period, the node occupying the channel transmits the preamble, receives the acknowledgment, and transmits the data frame immediately after any backoff. this is

to be. When a node successfully receives a data frame from Sched-wakeup, the energy consumption

Is as follows.

다음과 같다.A collision occurs when two or more TX nodes have the same receiver and select the same contention window size in Sched-wakeup. The energy consumption when a node does not successfully receive data in the sched-wakeup period

As follows.

기간 동안 데이터 프레임을 전송하거나 수신하지 않고 액티브 및 유휴 상태를 유지할 수있다. Sched-wakeup에서 유휴 상태를 유지하는 노드의 에너지 소비량

은 다음과 같다.Finally,

It can remain active and idle without transmitting or receiving data frames for a period of time. Energy consumption of nodes that remain idle in a sched-wakeup

Is as follows.

Sched _ wakeup
Synch_wakeup
Rx_Success
Rx_Failure
Idle
Tx_Success
Case 1
Case 2
Case 3
Tx_Failure
Case 4
Case 5
Case 6
Sleep
Case 7
Case 8
Case 9

Since all nodes of the RIX-MAC have two independent queues to transmit and receive in Sched-wakeup, the node operation in the period can be classified into nine cases as shown in [Table 1]. Case 1 means that the node successfully transmits data frames and receives data frames during Synch-wakeup and Sched-wakeup periods. Case 2 shows that one node successfully transmitted a data frame but did not receive a data frame in Sched-wakeup. In Case 3, the node successfully transmitted a data frame in Synch-wakeup and there is no data frame destined for the node Lt; / RTI > Similarly, another case represents the operation of the node in a similar manner. Thus, the node average energy consumption of the period can be obtained by multiplying the energy by the probability of all cases as shown in [Table 2].

Cases
Energy consumption

One

2

3

4

5

6

7

8

9

는 40 슬롯입니다. 프리앰블 프레임,

및 얼리 애크 프레임,

를 전송하는 지속 시간은 하나의 슬롯으로 설정되고, 데이터 프레임,

를 전송하는데 사용되는 시간은 5 슬롯이다. 노드와 네트워크의 다른 주요 매개 변수는 [표 3]에 나와 있다.Next, performance evaluation of RIX-MAC is presented in terms of throughput, delay and energy consumption of the network. To verify the proposed model of RIX-MAC and to evaluate the performance of RIX-MAC, we compare the numerical results of the proposed model with the simulation results using NS-2 and show the performance of RIX-MAC. Performance of X-MAC in terms of throughput and energy consumption. For analysis and simulation, it is assumed that the network is fully connected and each node has a single omnidirectional antenna in which the propagation model of bi-directional ground reflection is used. The queue capacity Q of each node is set to 10. The slot time τ is 1 ms, and the activation time

There are 40 slots. Preamble frame,

And early arc frame,

Is set to one slot, and the data frame,

Lt; RTI ID = 0.0 > 5 < / RTI > The other key parameters of the node and network are shown in Table 3.

Parameters

Notation
Value
Channel bandwidth

250 kbps
Active period in a cycle

40 ms
Time for a preamble frame

1 ms
Time for a ack frame

1 ms
Time for a data frame

5 ms
Power in Rx

15.2 mW
Power in Tx

28.9 mW
Power in sleep state

0.0004 mW

mW
Queue length
Q
10
Back-off window
W
32
Total simulation time

1,000 sec

과 각 노드가 경쟁을 이길 확률 p는 [수학식 7]. [수학식 8] 및 [수학식 11]을 풀고 모델의 정적 분포를 구함으로써 얻을 수 있다, π는 방정식 [수학식 1]에서 [수학식 6]에 p를 연결하여 계산할 수 있다. [수학식 13]을 사용하여 처리량을 얻는 데 사용되는 표 3의 다양한 듀티 사이클 비율 및 매개 변수 아래의 [10]에서 설명한대로 한 쌍의 확률 (p,

). 예를 들어 듀티 사이클이 14.3% 이고, 12 노드가 네트워크에 있고 백 오프 윈도우가 32 인 경우 p는 0.77로,

은 0.42가 된다.Static probability of empty queue state

And p is the probability that each node will win the competition. Can be obtained by solving [Equation 8] and [Equation 11] and obtaining the static distribution of the model. [Pi] can be calculated by connecting p to Equation 6 in Equation [1]. As described in [10] under the various duty cycle ratios and parameters in Table 3 used to obtain throughput using Equation (13), a pair of probabilities (p,

). For example, if the duty cycle is 14.3%, 12 nodes are in the network, and the backoff window is 32, p is 0.77,

Becomes 0.42.

FIGS. 6 to 14 show the results of simulation of the RIX-MAC protocol and the X-MAC protocol according to the present invention under various circumstances.

The dotted line represents the performance result of the X-MAC, and the solid line represents the performance RIX-MAC result. Fig. 6 shows the average throughput, Fig. 7 shows the average delay, Fig. 8 shows the total energy consumption according to the data arrival rate, the cycle length and the number of network nodes.

As shown in FIG. 6, as the number of nodes increases, the delay of the X-MAC linearly increases but the delay of the RIX-MAC increases slightly. If there are more nodes in a network using X-MAC, the chance of each node transmitting frames within the period is reduced, and the probability of long queues is increased, so that the delay time becomes longer.

As shown in FIG. 7, the throughput per node of the X-MAC and RIX-MAC is not affected by the number of nodes in the network. This is because successful transmissions increase with the number of transmission failures (collisions) as the number of nodes. However, RIX-MAC outperforms X-MAC in terms of throughput per node. Since the X-MAC wastes energy to transmit a short preamble in the case of collision, as shown in FIG. 8, when the number of nodes of the network increases, the energy consumption per node increases per cycle in the X-MAC. However, the RIX-MAC can save energy because the node of the RIX-MAC does not access the channel, consumes energy, and does not transmit the preamble.

When the cycle length is small, as shown in Fig. 9, all the reception frames can be transmitted with a low waiting time after arriving at the network. However, as cycle length increases, saturation and queue overflows increase contention delay and queue delay.

As shown in Fig. 10, the throughput of the RIX-MAC changes little by little as the cycle length increases. However, the throughput of the X-MAC linearly decreases as the period length increases due to the low duty cycle and the short active period carrying the incoming frame. If the length of the cycle is long, it is assumed that at most one frame is transmitted in one cycle, so the throughput is low.

As shown in FIG. 11, the energy consumption per node during the period maintains a constant in the RIX-MAC regardless of the period length since all nodes that are not involved in frame delivery can sleep for a longer period of time.

As shown in FIG. 12, in the X-MAC, the delay of the RIX-MAC initially increases almost linearly as the arrival rate increases with the relatively low arrival speed, but the arrival rate The higher the value, the higher the linearity does not increase. The delay of the RIX-MAC does not increase as the arrival rate increases because the queue of each node is not filled with frames with this arrival rate.

As shown in FIG. 14, as the load of the incoming frame increases, the energy consumption per node increases because more energy is required to transmit more frames. However, when the RIX-MAC becomes saturated and the queue for each node overflows, more frames are dropped by the queue and the number of frames to be transmitted remains the same. Thus, the energy consumption of the RIX-MAC is limited as the load of the received frame increases further.

 Thus, the RIX-MAC protocol according to the present invention is designed to minimize energy consumption by using a short preamble and allowing the sender to predict the recipient's wake-up. Transmission collisions can occur over wireless channels. In particular, burst traffic, which is common in sensor networks, can occur. The backoff mechanism of RIX-MAC also reduces the likelihood of transmission collisions for multiple senders.

The RIX-MAC protocol is superior to the X-MAC protocol in terms of latency, throughput and energy consumption by avoiding collisions and reducing overhead. As a result, the RIX-MAC can perform rapid communication in a wireless sensor network without sacrificing energy savings at both the transmitting and receiving nodes, and can reduce the collision and unnecessary wake-up, thereby extending the service life of the sensor network.

Meanwhile, the RIX-MAC protocol and the wireless communication method using the RIX-MAC protocol according to the present invention are not limited to the configuration and method of the embodiments described above, but the embodiments may be implemented by various implementations All or some of the examples may be selectively combined.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments, but, on the contrary, It should be understood that various modifications may be made by those skilled in the art without departing from the spirit and scope of the present invention.

Claims (8)

In a wireless sensor network, a first node is woken up to generate a preamble inserted with a destination address, and repeatedly transmitting the preamble at an interval capable of receiving Early-ACK;
Receiving a preamble at a second node and transmitting an early acknowledge (ACK) message including time information for predicting its own wakeup schedule to the first node when the destination address and its own address match; And
And transmitting data from the first node to the second node. The method according to claim 1,
Wherein the first node refers to the time information and wakes up according to the wakeup schedule of the second node. The method according to claim 1,
Wherein the time information is information on how much time has elapsed since the second node wakes up and transmitted the early acknowledgment. The method according to claim 1,
Further comprising switching to a sleep state after the first node completes data transmission to the second node. The method according to claim 1,
The first node not receiving the early acknowledgment from the second node within a predetermined time, and switching to a sleep state. The method according to claim 1,
Further comprising performing a random backoff procedure to prevent the first node from colliding with another node in preamble. 1. A wireless sensor network, comprising: a first node that repeatedly transmits a preamble while generating a preamble in which a destination address is inserted in a wake-up state so as to receive an early-acknowledgment (ACK); And
And a second node for receiving the preamble and transmitting an early acknowledgment message including time information for predicting its own wakeup schedule to the first node when the destination address matches its own address,
Wherein the first node transmits data to the second node after receiving the early acknowledgment, and refers to the time information to adjust the next wakeup to the wakeup schedule of the second node system. 8. The method of claim 7,
And after the first node completes the data transmission to the second node, switches to the sleep state.

Images