On the recursive nature of end-to-end delay bound for heterogeneous wireless networks

Multi-hop wireless networks are increasingly becoming more relevant to current and emerging wireless network deployment. The need for understanding the performance of such networks in order to be able to provide quantifiable end-to-end quality of service is apparent. Until recently, only asymptotic results that describe the scaling of the delay in the size of the network under numerous conformity conditions were available. Recently, a new methodology for wireless networks performance analysis based on stochastic network calculus was presented. This methodology enables the computation of end-to-end probabilistic delay bound of multi-hop wireless networks in terms of the underlying fading channelparameters. However, the approach assumes identically distributed channel gain which applies to a very specific class of networks. In this work, we seek to develop an end-to-end probabilistic delay bound for multi-hop wireless networks with non-identically distributed channel gains. We show that the delay bound for such networks can be computed recursively. We validate the resulting bound by means of simulation and discuss various numerical examples.