Research On MAC Protocol In Industrial Wireless Sensor Networks

With development of communication and hardware technology, because of the self-organization, flexible expansion and other technical advantages, wireless sensor networks are expected to offer promising solutions for industrial system in the wireless sensing and wireless control. However, complex usage of industrial wireless technologies in the harsh industrial environment has induced serious interferences on wireless sensor networks, the fact of which brings a great variety of challenges to the performances on reliability and real-time in the application-oriented industrial wireless sensor networks.Regarding the key technologies of MAC layer of industrial wireless sensor networks, the main contributions of this thesis are as follows.Based on the industrial environment, we intend to describe the system architecture of the industrial wireless sensor networks in this thesis, and we analyze the impact of industrial wireless channel on the wireless signal transmission through the actual testing.In this thesis, a multi-user medium access protocol (named IWSN-MAC) suitable for industrial wireless sensor networks is proposed. In traditional wireless sensor net-works, both contention-based CSMA protocol and scheduling-based TDMA protocol are unable to meet the requirements for industrial real-time applications. In this thesis, re-garding different time-constraints of industrial tasks, we predefine a set of event priorities through the method of pre-assigning orthogonal frequencies, which are taken into ac-count different sub-carriers for requiring channel access to the centralized device. Based on event priorities, a novel judgment mechanism used to determine the order of accessing channel is proposed. And in according to the complexity of industrial environment and the randomness of competition in the wireless channel, the centralized device decides the final channel access priority order based on event priories, and device can access channel in each transmission phase one by one in according to the final decision. In or-der to investigate the performance of IWSN-MAC protocol, the stochastic behavior of devices to transfer data is analyzed through DTMC modeling, we estimate the perfor-mance on throughput and delay. The analysis results show that, as an on-demand MAC protocol which can be used in dynamic network scenario, IWSN-MAC contributes to not only avoid the uncertainty delay arising from channel conflict, but also ensure deter-ministic and predictable real-time for industrial wireless sensor networks. In addition, IWSN-MAC especially adapts to the time-critical applications. It can also strengthen performance on channel utilization.A centralized channel allocation algorithm (named TS-CA) which is suitable for industrial wireless sensor networks is proposed. Industrial wireless sensor network is merged from traditional wireless sensor networks and other networks, because of the constraint of the wireless channel resource, channel multiplexing easily produces data collisions. To avoid the interferences, from the perspective of artificial intelligence theory, TS-CA which is based on tabu search heuristic scheme can be used to provide a globally optimal channel allocation solution through the continuous iterative search process. TS-CA can provide a better channel subset for each device in the network. In order to avoid the endless looping problem on the iterative process of searching local optimal solution, the interference temperature is taken account into being the condition to terminate TS-CA algorithm. The analysis results show that TS-CA algorithm is effective due to channel interference’s avoidance in industrial wireless sensor networks.An adaptive power control algorithm (named CRN-PC) which is also suitable for in-dustrial wireless sensor networks is proposed in this thesis. Based on the characteristics of industrial wireless sensor networks, from the perspective of cognitive radio network, we respectively model cumulative interference for primary system and secondary system in according to interference temperature. And then, the problem which targets at ensuring primary system reliability and ensuring secondary system reliability simultaneously can be modelled to be an optimization problem, the objective of which is to maximize the net-work utility with a couple of anti-interference constraints. This part applies utility theory into power control in the industrial wireless sensor networks. According to mathematical programming theory, we give a solution. The analysis results show that the ultimately equilibrium point of CRN-PC is the expected optimal transmission power targeted for ensuring communication reliability. Furthermore, the optimal value is proved to be the globally optimal transmission power based on the convex optimization theory. It can also strengthen performance on power utilization.