Browse Prior Art Database

CENTRALLY-CONTROLLED ADAPTIVE JITTER BUFFER FOR RF SITES

IP.com Disclosure Number: IPCOM000125463D
Original Publication Date: 2005-Jun-01
Included in the Prior Art Database: 2005-Jun-01
Document File: 4 page(s) / 46K

Publishing Venue

Motorola

Related People

Robert J. Armstrong, Jr: AUTHOR [+2]

Abstract

IP (Internet Protocol) networks offer some considerable advantages over circuit-switched networks for transporting control, voice, and data traffic across a wide area system, but these advantages do come at a cost. IP networks are not as reliable as their circuit-switched counterparts; packet delivery is not guaranteed, and individual packets can be delayed differently as they pass through the IP network. This phenomenon is known as “jitter.” Jitter is especially troublesome for wireless communication systems, since transmitters receiving packets off the IP network must have continuous data to send out over the air, lest holes may appear between critical bits of information. This starvation of the transmitter degrades the performance of the system.

This text was extracted from a Microsoft Word document.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 53% of the total text.

CENTRALLY-CONTROLLED ADAPTIVE

JITTER BUFFER FOR RF SITES

By Robert J. Armstrong, Jr. and Paul G. Bell

 
 

IP (Internet Protocol) networks offer some considerable advantages over circuit-switched networks for transporting control, voice, and data traffic across a wide area system, but these advantages do come at a cost. IP networks are not as reliable as their circuit-switched counterparts; packet delivery is not guaranteed, and individual packets can be delayed differently as they pass through the IP network. This phenomenon is known as “jitter.” Jitter is especially troublesome for wireless communication systems, since transmitters receiving packets off the IP network must have continuous data to send out over the air, lest holes may appear between critical bits of information. This starvation of the transmitter degrades the performance of the system. 

Many IP-based wireless communications systems employ a buffer at their transmitters in order to protect against network jitter. This jitter buffer delays the transmission of the first packet received from the network so that subsequent packets can arrive at the destination site with sufficient time to go out over the air without any breaks in-between. Traditional jitter buffers are typically static, in that they implement a hard-coded timer which is of sufficient size to handle the expected worst-case network conditions. While this static implementation is simple, it is not necessarily efficient under all wireless transaction scenarios – it may be insufficient under some conditions or scenarios, while it may be excessive for others. Wireless systems which transmit messages with different packet sizes and/or inter-packet spacings are particularly prone to these inefficiencies.  A centrally-controlled jitter buffer which is adapted on a per-transaction basis can be used to improve system performance (by ensuring that there is a sufficient amount of buffer) and to improve system throughput (by ensuring that the buffer is not excessive).

Different audio or data sources in a system may launch packets onto the network at different intervals. One source might launch packets 20ms apart, while another launches them 120ms apart. This has a very significant and direct impact on the jitter buffer size at the packets’ destination. Unfortunately, a destination which employs a static jitter buffer is not aware of the packet-launching behavior of the source and cannot adjust its buffer size based on this behavior. In addition, a network might be laid out in such a manner that some areas of the network might be more prone to large amounts of jitter than other areas. A destination site is unaware of this special circumstance and cannot adjust the size of its jitter buffer to compensate. A central controller which knows the types of sites in the system and contains information about the network topology can be used to instruct destination RF sites involved with a transaction to increase or decrease their j...