This paper presents a novel architecture for providing converged IP-based TV (IPTV) services specified by ETSI TISPAN standardisation for IPTV in ongoing NGN release 2 specifications. The described IPTV architecture is based on utilisation of the IP Multimedia Subsystem concept used by NGN architectural framework and its adaptation to provide the IPTV specific functionalities and services. Using the foundation provided by the IMS based architecture, we propose a new functional architecture to enhance the functionalities and features needed for scalable converged networks, flexible media delivery and advanced IPTV service scenarios. The proposed architecture, leveraging on the FMC architecture that operators may deploy to provide IPTV service across different access networks in future deployments (mobile, wireless, fixed) has prototypically been implemented in the ScaleNet* demonstrator testbed. This paper analyses in detail the main principles for such a converged reference architecture. The paper also presents the IPTV service scenario prototype called Click-to-Multimedia which shows some basic features and advantages implemented on top of the presented architecture by prototyping demo applications as proof of concept reference.

No abstract

No abstract

Wireless mesh networks (WMNs) are wireless multihop backhaul networks in which mesh routers relay traffic on behalf of clients or other routers. Due to large MAC layer overhead, applications such as Voice over IP, which send many small packets, show poor performance in WMNs. Packet aggregation increases the capacity of IEEE 802.11-based WMNs by aggregating small packets into larger ones and thereby reducing overhead. In order to have enough packets to aggregate, packets need to be delayed and buffered. Current aggregation mechanisms use fixed buffer delays or do not take into account the delay characteristics of the saturated IEEE 802.11 MAC layer. In this work, we present FUZPAG, a novel packet aggregation architecture for IEEE 802.11-based wireless mesh networks. It uses fuzzy control to determine the optimum aggregation buffer delay under the current channel utilization. By cooperation among neighboring nodes FUZPAG distributes the buffer delay in a fair way. We implemented and evaluated the system in a wireless mesh testbed. For different network topologies we show that FUZPAG outperforms standard aggregation in terms of endto-end latency under a wide range of traffic patterns.

Citation: Sivchenko, D., Rakocevic, V. & Habermann, J. (2015). Integrated mobility and resource management for cross-network resource sharing in heterogeneous wireless networks using traffic offload policies. Wireless Networks, 21(3), pp. 981-999. doi: 10.1007/s11276-014-0826-7 This is the accepted version of the paper.This version of the publication may differ from the final published version. Abstract: The problem of efficient use of resources in wireless access networks becomes critical today with users expecting continuous high-speed network access. While access network capacity continues to increase, simultaneous operation of multiple wireless access networks presents an opportunity to increase the data rates available to end-users even further using intelligent cross-network resource sharing. This paper introduces a new Integrated Mobility and Resource Management (IMRM) framework for automatic execution of policies for cross-network resource sharing using traffic offload and pre-emptive resource reservation algorithms. The presented framework enables both mobile-initiated and network-initiated resource sharing policies to be executed. This paper presents the framework in detail and analyses its performance using extensive ns-2 simulations of the operation of a set of static policies based on measured signal strength, and dynamic pre-emptive network-initiated policies in a WiFi/WiMAX scenario. The detailed evaluation of the static policies clearly shows that the quality of voice applications shows large deviation, mostly due to very different levels of delay in access networks. Based on these conclusions, this paper presents a design of two new dynamic policies and shows that such policies, when efficiently implemented using the new IMRM framework can greatly improve the capacity of the network to serve voice traffic with a minimal impact on the data traffic and with a very low signalling overhead. Permanent