Chip improves performance of 3G networks

Bell Labs’ new codec will increase the efficiency of third-generation universal mobile telecommunications systems.

Bell Labs, the research and development arm of Lucent Technologies, has developed a new codec that will increase the efficiency of the company’s third-generation (3G) Universal Mobile Telecommunications System (UMTS) network equipment.

A combined team of researchers from Bell Labs Australia and Lucent’s Mobility Solutions Group developed the codec at the company’s North Ryde facility. They then worked with development teams in France, the UK and the US to incorporate the technology into silicon chips that are being included in Lucent’s Flexent OneBTS base station for UMTS operators.

Channel decoding is the function in wireless base stations and handsets that typically requires the most computing resources.

The Bell Labs innovation, described as a ‘Unified Turbo/Viterbi Channel Decoder,’ is claimed to be the only available silicon codec implementation with built-in support for 3GPP turbo interleaving algorithms, bit-error-estimation for accurate power control and first level 3GPP channel de-interleaving.

These features enable the decoder to support many more voice and data calls than competitors’ solutions, which typically require software intervention to change service combinations.

The decoder supports any mix of voice and data services needed for UMTS networks and is currently the only channel decoder that supports both voice and data services for any 3G digital wireless standard.

A key benefit of the Bell Labs core design is the use of a patented turbo decoding termination algorithm, which yields low power consumption for channel decoding, a key consideration for handset manufacturers. Bell Labs also secured additional power consumption reductions by combining voice and data channel decoding functions, thereby yielding much smaller chips.

In fact, tests have demonstrated that the Unified Turbo/Viterbi Channel Decoder can decode UMTS 2 Mbit/sec data streams with 10 turbo iterations using a log-MAP algorithm with variable width log-correction tables. The decoder provides up to a 12% increase in link sensitivity over other available max-log-MAP solutions.

On the web