Low cost energy-efficient (power based) routing protocols of mobile ad hoc networks (MANETs) increase the lifetime of static networks by using received signal strength (RSS) and battery power status (PS). They require GPS service to find the exact location of mobile nodes. The GPS devices themselves consume power because they need excessive updates to find the stationary nodes for efficient routing. To overcome this, RSS is being used as a metric, followed by, residual battery power. The recent protocols, based on these concepts, provide energy efficient routes during the route discovery phase only. Topological changes make these routes weak in due course of time. To update routes, HELLO process can be used, which however creates unnecessary overhead, delay and consumes power. Hence, these protocols do not update the routes. We propose an energy-efficient reactive routing protocol that uses the RSS and PS of mobile nodes. Proposed Link Failure Prediction (LFP) algorithm uses the link-layer feedback system to update active routes. We use ns2 for simulation of the proposed algorithm. Comparing the results of proposed scheme and existing scheme, in terms of energy consumption, link failure probability, and retransmission of packets, we observe that the proposed scheme outperforms the existing one.

Replication strategy is mostly useful for replicating the called mobile user's profile at selected locations where some caller mobile users have higher calling rate. It plays important role in personal communication services (PCS) in order to reduce call setup cost and call setup time. In prior replication schemes, replicated sites (visitor location registers) have no inter relationship. We have proposed a novel and innovative replication tree which is buildup from replicated VLRs (visitor location registers). VLRs are replicated based on effective replication probability of caller mobile user. Home location register (HLR) keeps update replica at root VLR in the replication tree and root VLR forwards update replica to a VLR in the replication tree on demand. The main objective of the proposed scheme is to reduce call setup time and call setup cost by enabling replication of called mobile user's location profile at certain visitor location registers (VLRs) where caller mobile user roams. We have presented analytical modeling which shows that setting inter-relationship among the replicated sites helps in further reducing the call setup cost.