Research And Implementation Of Cross-layer Protocol For Wireless Sensor Networks

As an emerging cutting-edge technology, Wireless sensor netwoeks attract international attention from academic and industrial. To meet the increasing demand, wireless sensor netwoeks have been widely applied in both civil and military fields. Industry automation is one of the. Network controller is in charge of highly automatic factory, therefore it should feed back instructions to the sensors in different parts of a assembly line according to the montoring data collectted by them. Traditional network protocols is designed for a network composed of same types of nodes, however, nodes distributed in a automatic factory always is likely to have differentiated requirements in communications. How to meet the requirements of differentiated nodes and how to put forward a a suitable protocol architecture have become a new challenge to the existing design approach.TinyOS is a lightweight embedded operating system for wireless sensor netwoeks applications, and combines lots of merits including open-source, modular programming, therefore it is easy to understand why our project selects TinyOS as the development tool. Based on the theoretical analysis of beacon-enabled PANs, the over-all structure of the MAC implementation is built. The MAC implementation is composed of several components which realize different core functions of MAC sublayer, it uses timer to set the beginning and the end of beacon-transmit/synchronization period, CAP, CFP, and it could switch periods with the help of Superframe-Phase interface.Analyzing heterogeneous nodes existing in industial wireless networked control systems, the article divides them into three functional type, WNC(Wireless Networked Control), A-IS(Advanced Industrial Sensor), S-IS(Simplified Industrial Sensor). In consideration of their differentiated requirements, a BACnet application layer is designed, which defines objects and services for every functional type.Based on the research of IEEE 802.15.4 standard and BACnet protocol, a new cross-layer optimization protocol is proposed for heterogeneous industrial wireless control networks, which optimizes data link layer in terms of the related information of the BACnet application layer. The protocol is composed of channel access management algorithm and distributed BO adaptive algorithm.In combination of BACnet protocol and IEEE 802.15.4 MAC sublayer standard, cross-layer protocol gives full play to the modular architecture and the event-driven nature of TinyOS. As the intermediate between application layer and data link layer, cross-layer information module inquires key information from BACnet application layer including device functional type, BACnet svcId, expected performance index, and then it decides channel access mechanism that data link layer should employ. Following different procedure that corresponds to CSMA/CA or GTS channel access method under the instruction from cross-layer information module, channel access management module organizes and manages MAC sublayer so as to complete transmission task. Dynamic GTS deallocation algorithm ensures that nodes with real-time requiremet have priority to be allocated Guaranteed Time Slot(GTS), meanwhile trying its best to make GTS in the superframe structure utilized effectively. In order to adjust the duty circle of end devices adaptively, distributed BO adaptive algorithm enables end device to tune its beacon order independently according to data transmission activity level without customized support from PAN coordinator. Furthermore, the algorithm is optimized for periodic transmission, if sampling period truly exists, local beacon interval would be adapted automatically to the value of possibe beacon interval thst is close to local beacon interval.At the end of the thesis, it describes how the protocol stack was evaluated based on the hardware platform equipped with CC2420 RF transceiver. The experimental results verified the effectiveness of the proposed channel access management module and distributed BO adaptive algorithm, proving that the project have realized design objectives as planned. Due to the complex environment of industrial automation networks, Further modification is still needed for the architecture of protocol and the details of algorithm according to the practical performance of engineering application.