Browse Prior Art Database

Joint Radio Resource and Mobile Battery Life Optimization in Cellular Data Networks Using Real-time Application Profiling Disclosure Number: IPCOM000236961D
Publication Date: 2014-May-23
Document File: 6 page(s) / 136K

Publishing Venue

The Prior Art Database


Signalling traffic in cellular data networks are increasing at very high rate due inefficient idle-to-active transition methods employed in current 3GPP standards. The user data traffic that flows through cellular network have a varied characteristics interms of packet sizes, flow duration and inter packet time. For example, there are very long background flows which have high interpacket times as well as there are short interactive flows that send bursts of packets. Current 3GPP standard proposes to use a fixed dormancy timer to determine when to transition a mobile's radio state from idle to active. This scheme is inefficient when there is wide variety of traffic characteristics. In this features, we propose a real-time traffic classifcation based radio resource optimization where the user traffic traversing through the network is characterized in real-time and the dormancy timer of the mobile is set in such a way that optimizes the number signaling transitions as well as the battery life of the mobile.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 26% of the total text.

Page 01 of 6

Joint Radio Resource and Mobile Battery Life Optimization in Cellular Data Networks Using Real-time Application Profiling

Cellular Data Networks (2G/3G/LTE) are becoming increasingly popular as the de facto data network for mobile communication as increasing number of smart phones, tablets and mobile devices connect to the Internet using them on the go. As more number of users connect to a mobile network, the load in the network increases. It has been predicted that the growth in mobile network load will be exponentially high in the next few years [1]. Traditionally, the load in the cellular data network is measured in terms of the number of bytes transferred through the network. But another significant component of load in a cellular network is the associated signaling traffic that is used to setup the radio bearers that carry the user traffic. This is due to the fact that a radio bearer needs to be established by the RAN (Radio Access Network) to transfer any user data and is then released when there is no more data sent from or to the user. Since Internet access is mostly chatty in nature, this causes tremendous signaling overhead to setup radio bearers and release them frequently.

The rise in signaling load in the network is causing significant scaling issues for the RAN in several cellular data networks worldwide [2, 3]. Service providers find their RNCs and eNodeB get overloaded significantly due to high signaling traffic. They have to either upgrade the RAN with high capacity machines or increase the number of RAN components to support the high signaling load incurring higher CAPEX and OPEX.

The problem we try to solve here, is to optimize the setup and release of radio bearers by analyzing user traffic profiles and there by reducing the signaling load in the network. Traditionally, voice traffic has a well defined access pattern such as a call setup phase, a period of constant bit rate voice traffic and call end phase. This works well with the creation and deletion of radio bearers. But data traffic have a much varied access patterns unlike voice as numerous types of applications such as video, chats, internet browsing, social networks, advertisements, analytics traffic, interactive games, news alerts etc access the Internet through the cellular data network. Our features profiles this application traffic of a user in real-time and optimizes radio bearer management, there by reducing the signaling load in the network.

There are several works in current literature that measure the impact of signaling overhead in the network but do not solve the problem [4, 5]. The 3GPP standard proposes fast dormancy where a mobile can let the network know when to release a radio bearer [6]. The draw back of this approach is more signaling in some cases where the mobile has very chatty traffic. Recent works in [7 8], take a mobile centric approach by optimizing mobile application traffic and there by indirectly managing the radio bearers i...