Network information theory for wireless communications
Liang‐Liang Xie, P. R. Kumar · 2003
Two fundamental questions arise con- cerning the emerging field of wireless networks. First, how much information can be carried over a wireless network with a multiplicity of nodes? Second, how should one operate wireless networks to realize this capacity? We establish that under a high attenua- tion regime, wireless networks obey a linear scaling law for the transport capacity, and the order opti- mal strategy is to relay packets from node to node after fully decoding them at each hop. This attenua- tion regime may be what normally prevails. There is currently much activity in the protocol development community in realizing this multi-hop mode of op- eration. On the other hand, when the attenuation is very low, superlinear scaling is feasible, and a strategy of coherent cooperation by upstream nodes to relay a packet to the next downstream node, along with decoding by interference cancellation, can be order optimal. These network information theoretic results provide a bridge between information theory and pro- tocol development.