Upper and lower bound on signal-to-noise ratio gains for cooperative relay networks

Cooperative networking as a means of creating spatial diversity is used in order to mitigate the adverse effect of fading in a wireless channel and increase reliability of communications. We investigate signal-to-noise ratio (SNR) gain in wireless cooperative networks. We show that the differential SNR gain in the high data rate regime, which we refer to as SNR gain exponent zeta_infin, is independent of the relaying strategy and only depends on the number of transmission phases used for communication. Furthermore, a straight-line upper and lower bound is derived based on geometric considerations. It is shown that the approximation error of the upper bound with respect to the exact SNR gain tends to zero for R rarr infin. For the lower bound, the approximation error tends asymptotically to a constant factor delta for R rarr infin. Both bounds are the best possible straight-line bounds with respect to absolute error.

I. Introduction

The randomness of the wireless channel leads to time-variant fluctuations of the received signal amplitude. This adverse effect of fading can be mitigated by diversity techniques [1], [2]. The creation of (spatial) diversity is one of the benefits of relaying and cooperation. In cooperative networks mobile nodes pool their resources in order to achieve a better performance, i.e., increase the reliability of communications. Along with those benefits come challenges for the design of large scale distributed networks such as cooperative relay networks. Those challenges can be settled in the field of network information theory (e.g., network coding) and communications theory (e.g., medium access, combining receiver structures).