| With the rapid development of wireless sensor networks (WSN) technology, the demanding of WSN's data transfer performance has been increasingly high. Currently, the most widely used single-channel MAC protocols are gradually eliminated, due to the low performance of network throughput caused by their own characteristics of low utilization of the channel resources. Multi-channel MAC protocol for wireless sensor networks can improve network transmission delay, packet loss rate and network throughput performance to a large extent, and can also solve the problem of network paralysis caused by interference. Whether the multi-channel MAC protocol for wireless sensor networks can coordinate the equitable use of the radio channel resources of sensor nodes, to safeguard the quality of the communication channel of service (QOS), can improve network throughput and capacity, reducing the transmission delay of the network, utilizing the limited network bandwidth to the largest extend or not has become the current research focus. Through the analysis of the existing wireless sensor networks shows that the existing protocol can't meet people's need of data transmission in many cases, take the hardware cost into consideration, therefore it means a lot to propose an efficient single transceiver multi-channel MAC protocol for wireless sensor networks. Through analyzing the problem that the multi-channel MAC protocols for wireless sensor networks need to solve, this paper proposed a single transceiver multi-channel MAC protocol based on IEEE802.15.4 protocolThe main research work of this project include:to solve the ubiquitous problem of serious packet loss rate and low data throughput caused by the uneven distribution of the channel load, this article proposed a load equalization channel allocation strategy based on ant colony algorithm, according to the advantage of its heuristic search, it approached the most balanced load distribution by adjusting the validate, residual and inspired factors. To solve the problem of high network delay caused by serious interference between different channels, this article design a channel quality prediction mechanism based on BP neural network, it adopt the measures that through several times self-learning of BP neural network to filter out the noisy channel. To solve the problem of the broadcast inefficiencies caused by excessive accumulation of broadcast packets, this paper proposed a broadcast mechanism based on the minimum network delay spanning tree, it formed the broadcast tree based on the minimum network delay to reduce the latency under the promise of broadcast coverage. To solve the inefficiency of the channel negotiation due to the inadequacy of the time synchronization, this article applied two-way packet exchange and parameter fitting combined method, through using least square linear fitting way to get more precise time synchronization equation, and then it will achieve more accurate entire network time synchronization. To solve the problem of long channel access delay, this paper designed an adaptive CSMA/CA channel access mechanism by adaptive ly adjusting the size of back off index and the length of the contention window to reduce the channel access latency. We redesigned the corresponding frame according to the characteristics of the MAC protocol we designed.Through the NS2 experiment, it indicated that the MAC we designed protocol performed better than the multi-channel MMAC protocol and IEEE802.15.4 protocol in terms of network throughput, network delay and packet loss rate to a large degree. Through the relevant hardware experiment, it proved that compared with traditional IEEE802.15.4 protocol, the single transceiver multi-channel MAC protocol designed in this paper, can effectively reduce the network delay, and improve network throughput and anti-jamming capability. Thus, it proved that the single transceiver multi-channel MAC protocol designed in this paper can meet the requirements of practical application. |
PDF Full Text Request