The main reasons for introducing cross-layer design to wireless networks can be illustrated as follows.
First of all, the assumptions in wireless networks are different from those of the wired networks . For instance, in wired networks, the packet losses are assumed to be a result of network congestion, while in wireless networks, packet losses are often caused by corruption. If TCP still invokes congestion avoidance mechanisms to deal with packet loss in wireless environments, it will make things worse. Explicitly notifying the packet corruption rather than congestion in the signaling from link layer to transport layer will resolve this problem efficiently .
Second, due to the resource constraints in MANETs, it is imperative that the various protocols in the protocol stack collaborate to use the limited bandwidth efficiently and reduce power consumption. For instance, poor channel conditions will lead to frame retransmissions and delay at the link layer, which, in turn, leads to TCP retransmissions. To solve this problem, retransmission information can be communicated between the link layer and transport layer; power consumption will be decreased by increasing MAC sublayer retransmission to prevent TCP from retransmissions [18, 19].
Third, with the acceptance of MANETs and wireless mesh networks, WMNs QoS requirements such as acceptable packet loss rate and bounded ...