Performance Analysis Of Cooperative Communication Based On Stochastic Network Calculus

The mobile Internet era accelerates the development of mobile communicationtechnology, the application of the fourth generation (4G) mobile communicationtechnology is faster and faster, the4G network will give users faster Internet speeds andbetter service experience. In4G network, the physical layer will use some more advancedtechnology including cooperative communications technology. The basic idea ofcooperative communication is to establish multiple virtual transmission links between thesource node and the destination node to complete the transfer of information through thenode between auxiliary. Thoughtful performance researches should be done aboutcooperation communication. Current researches about cooperation communicationperformance mainly focus on the physical-layer such as the bit error rate, the outageprobability and MAC access. The performance of network layer which bears moreintuitive value such as delay and data backlog is rarely analyzed.Stochastic network calculus is a mathematical tool used to analyze wireless channelquality of service (QoS) and get the network delay and data backlog. It can solve thelimitations of the traditional tools, such as probability theory, stochastic processes, andqueuing theory. Stochastic network calculus can easily analyze wireless network delayprobability distribution and the data backlog by charactering the data flow and networkservice capabilities respectively. This work mainly focuses on modeling service ofcooperation communication system and random service characteristics in amplify-and-forward mode.Based on stochastic network calculus, we firstly model a three-node amplify-and-forward mode cooperative communication channel and get the stochastic service curve ofthe channel. Taking into account the self-similarity characteristic of the data flow inmodern communication network and get the stochastic arrival curve. And then we canderive the delay and backlog performance distribution of amplify-and-forward cooperativecommunication system. Finally, the numerical analysis of the performance distribution isdone and the effect of self-similarity on the network performance is analyzed.