No abstract

No abstract

No abstract

A tight coupling of LTE and Wi-Fi interfaces can be achieved by integrating them at their radio protocol stacks. LTE and Wi-Fi radio level integration with IPSec tunnel (LWIP) was introduced by 3GPP as part of Rel-13. This tighter level of interworking replaces the traditional way of cellular-Wi-Fi interworking through a packet gateway and it can react to the dynamic changes in the wireless link quality. In this paper, we present a variant of LWIP prototype that works with commercial UE (Nexus 5). The developed LWIP prototype uses OpenAirInterface (OAI) for LTE network and Cisco Access Point (AP) as Wi-Fi AP. We also present the design and implementation of LWIP prototype and interesting results for tight interworking of LTE and Wi-Fi at IP level. We have evaluated the LWIP performance with different Link Aggregation Strategies (LAS) using both UDP and TCP. We have observed that, in a highly loaded Wi-Fi channel, when LWIP employs Wi-Fi only in Downlink (WoD) LAS, then sum of individual TCP flow throughput has improved by 28% as compared to LWIP operating with Flow Split (FS) LAS. We have enumerated the challenges which has to be addressed in LWIP to reap the maximum benefits.

A wireless ad hoc network is typically Constrained by battery power, and hence, it is important to conserve battery power in order to keep the network alive for a long time. We consider a wireless network consisting of an access point (AP) and a wireless node, which is limited by energy, and we seek scheduling algorithms that conserve battery power at the wireless node. Since the packet arrivals are stochastic, communication happens only during the time slots when packets are available. Hence, keeping the receiver ON at all times is wasteful of resources. However, when the receiver is turned OFF, a packet, on arrival, is stored in a buffer (at the AP) which incurs a holding cost. We propose sleep/wake scheduling at the wireless node with an objective of minimizing the power spent (at the receive node) and the queuing delay at the AP, while maintaining the stability of the queue. We note here that this energy optimum scheduling is a distributed problem in which the queue-length information at the transmitter is not known at the receiver. We show that the optimum wake process at the receiver follows closely as that of the arrival process of the packets at the AP. We numerically evaluate the optimum scheduler, and discuss the energy and queuing delay performance of our proposed optimum scheduler.Index Terms-Energy efficiency, network lifetime, sleep/wake scheduling I.