Research On Transmission Reliability And Error Packet Recovery Algorithm Based On IEEE 802.15.4 Protocol

Wireless sensor networks(WSNs)are widely used in environmental monitoring,smart home,healthcare,smart factories and other fields due to their low energy consumption,low cost and wireless transmission.However,due to signal attenuation,multipath effects,and co-channel interference,wireless signals are easily affected,leading to packet loss.In the existing research,partial retransmission methods require additional energy,and forward error correction(FEC)technology adds redundant bytes to the data packet,which sacrifices network throughput and increases latency.Therefore,how to improve network performance without increasing energy consumption and reducing payload is still a challenging problem.The IEEE 802.15.4 protocol is the basis of many wireless technologies,such as: Zig Bee,Wireless HART,and so on.Based on the transmission characteristics of the pseudo-noise code in the physical layer,this paper the bit error patterns of the pseudonoise code.The three proposed packet recovery algorithms significantly improve the packet reception rate.Firstly,we establish a wireless network through Zig Bee nodes and Universal Software Radio Peripheral(USRP).Through extentive experiments,we explored the bit error characteristics of original pseudo-noise code received in the physical layer of the IEEE 802.15.4 protocol.Depending on how much the data packet was corrupted,we divided the received packets into four categories: nonerror packet,error-tolerable packet,discrete-error packet and continuous-error packet.In those packets containing erroneous pseudo-noise code,we found three significant bit error patterns: Bit LagOffset Error,Bit Advance-Offset Error and Bit Mutation Error.We further analyze these bit error patterns in different error packets.Secondly,based on these three significant bit error patterns,we propose two error packet recovery algorithms: offline recovery algorithm OffER and online recovery algorithm On ER,which can directly recover erroneous packets without retransmission or adding redundant bytes to the packets.We have evaluated the effects of these two algorithms on outdoor open lawns and indoor simulated industrial environments,and have significantly improved the packet reception rate.Finally,because the environmental impact on wireless signals is very difficult to model and analyze.We propose an error packet recovery algorithm named NNER based on neural network to decode the pseudo-noise code,including pseudo-noise codes in those error packets.The experimental results on outdoor open grass and indoor simulated industrial environment show that NNER can significantly improve the packet reception rate.