1x Evolution-Data Optimized Revision A (1xEV-DO Rev. A) is a cellular communications standard that introduces key enhancements to the high data rate packet switched 1xEV-DO Release 0 standard. The enhancements are driven by the increasing demand on some applications that are delay sensitive and require symmetric data rates on the uplink and the downlink. Some examples of such applications being video telephony and voice over internet protocol (VoIP). The handoff operation is critical for delay sensitive applications because the mobile station (MS) is not supposed to lose service for long periods of time. Therefore seamless server selection is used in Rev. A systems. This research analyzes the performance of this handoff technique. A theoretical approach is presented to calculate the slot error probability (SEP). The approach enables evaluating the effects of filtering, hysteresis as well as the system introduced delay to handoff execution. Unlike previous works, the model presented in this thesis considers multiple base stations (BS) and accounts for correlation of shadow fading affecting different signal powers received from different BSs. The theoretical results are then verified over ranges of parameters of practical interest using simulations, which are also used to evaluate the packet error rate (PER) and the number of handoffs per second. Results show that the SEP gives a good indication about the PER. Results also show that when considering practical handoff delays, moderately large filter constants are more efficient than smaller ones.

In mobile IP, a triangular routing problem usually leads to additional delays and non-seamless handoff, which causes the loss of a large amount of in-fly packets. In this thesis we propose an adaptive route optimization scheme that considers the seamless handoff and mobility rate. The proposed scheme adopts the mobile routing table scheme and includes an efficient cost function. The cost function is a trade-off between network bandwidth utilized by the routing path, signaling messages and processing loads of agents. The primary idea is to adaptively determine when to perform route optimization. The effect of the mobility rate and the cell sojourn time of a mobile node are studied. Both new call and handoff call are considered in the cost calculation. Simulation result shows that the proposed scheme can effectively reduce the total cost composed of link and signaling costs.