Scalable service guarantees through stochastic modeling for bandwidth aggregation of multiple wireless links during mobility

The E2E delay constrained multi-path flows in heterogeneous wireless networks face major challenge of out-of-sequence (OOS) reception of packet due to asymmetric path characteristics. The larger spread of OOS reception generally leads to expensive buffer handling for in-order delivery of packets. The problem gets adverse during mobility, as the change in path due to change of point-of-attachment (PoA) cause unpredictable E2E delay. In this paper, the multi-server scheduling to minimize OOS reception has been investigated with the help of novel stochastic model that evolves tight upper-bounds for crucial metrics such as E2E delay distribution and accompanying OOS reception. The model provides useful formulae for the Quality-of-service (QoS) dimensioning during mobility. The paper evolved some approximate estimates of E2E delay variations and its impact on OOS reception of packets. The main application of the proposed model has been demonstrated through scheduling strategies that minimize E2E delay variation and OOS reception in multipath transport during mobility. The results of proposed model have proven the effectiveness in minimizing impact of asymmetric characteristics of heterogeneous networks.