The use of multiple antennas such as the examples in the scenarios SIMO, MISO, and MIMO is a key feature in the concept of cooperation. When cooperation needs to be implemented in a reconfigurable network, such diversity can be obtained by coordinating several transmitters or receivers in order to achieve the same advantages as those for the MIMO channel (i.e., diversity). Another important issue to consider when looking into cooperative schemes is that not only physical layer performance must be measured, but also the effects and advantages of the coordination at higher layers to improve performance in an integral way in the network architecture.
In order to implement cooperative communications in reconfigurable networks, nodes must be organized in some logical way so that they recognize potential helpers as a cluster. Such a group of nodes can be commanded by an AP that can be fixed or mobile but has more processing capability; or the group of nodes could be coordinated by the node in need for a certain period of time during which cooperation must take place.
In Scaglione et al. , a discussion of these fundamental principles is introduced, where BER and outage are chosen as performance metrics. They compare the noncooperative transmission versus the cooperative scenario, considering in principle a synchronous system with the amplify-and-forward strategy, as well as a comparison with the decode-and-forward strategy. It is shown that cooperation has advantages over noncooperative methods. Also, they discuss the issue of cooperation as a networking strategy to improve performance with influence and effects on lower layers, such that the concept of an end-to-end link remains to be well defined.
In Hong et al. , cooperative schemes with the objective of improving performance by relaying information through several nodes in the network and optimally allocating power and bandwidth are introduced. In order to achieve such cooperation, channel state information is used. Hong et al. also recommend defining strategies to consider the channel state information estimation improvement to achieve better power allocation.
Cooperation in multihop wireless networks is still an open issue to be studied and needs to be considered within the cross-layering framework together with the end-to-end link abstraction. Cooperative schemes increase the communication cost in the network, and strategies to decrease this cost as well as their robustness to time-varying characteristics, such as network topology, need to be investigated.
In general, nodes in the network must be able to decide by themselves whether to accept a relaying request. This would lead to a noncooperative or cooperative scheme depending on the decision taken by those nodes. The decision-making process must depend on variables such as environment and internal energy, since ultimately, it will determine the degree to which relaying can take place.
In Srinivasan et al. , a network with this idea of considering remaining energy in the nodes is introduced, and an algorithm to determine the optimal proportion of cooperation that each node should receive is formulated. The results show that such an algorithm converges to the optimal operation point. The remaining energy is considered by dividing the nodes in the network into classes where each class has an energy constraint and an expected lifetime. The results are presented by formulating a game theory framework for the algorithm.
The notion of topology organization is also helpful in establishing cooperative algorithms in reconfigurable networks. Nodes can be organized in clusters, and then each of those groups can determine which nodes will serve as intercluster communicators. In del Coso et al. , an algorithm that considers cooperative diversity using MIMO channels in wireless sensor networks is presented, time is slotted, and one of the time slots is used for internal communication in the clusters and another time slot is used for intercluster communication.
Energy constraints are also used in each of the links belonging to the end-to-end communication, and the algorithm achieves minimum outage probability in the entire trajectory by deriving the optimum times at which intra- or inter-cluster communication must take place at every hop as well as the power allocation in the route. The results show that diversity like that presented for MIMO channels is achieved using this algorithm.