TfI has available for licensing, a new and novel error correction algorithm for algebraically generating encoding and decoding matrices. It is a platform technology with many potential applications however it is particularly suited to low latency, low power consumption wireless applications  requiring  the  in-line  generation  of  LDPC  codes  with  pre-defined  parameters (size, rate, target BER).

Technology Overview

Error  correction  coding  algorithms  are  used  to  protect  data  from  losses which may  occur before  the data  is used  again. The principle  cause  for  such  losses  is due  to disturbances  in transmission  of  the  data  from  one  point  to  another.  Standard encoding algorithms add additional information bits (parity-check bits) to the transmitted data by using large matrices which are operated on the data to be encoded. To retrieve the data the encoded information is operated on by another matrix that allows for the recovery of lost information on the basis of the parity-check bits. Each potential matrix is rated on its ability to recover lost information (its 'distance') and on its ability  to minimise  the additional encoded data needed  to achieve this (its 'rate'). There is a fundamental limit of the amount of information that can be reliably transmitted  via  a  given  channel  -  the  so  called  'Shannon  limit'.  Two  types  of  codes  are currently  closest  to  the Shannon  limit  -  turbo  codes  and  low  density  parity  check  (LDPC) codes.

 Most  modern  communication  and  storage  systems  use  encoding  and  this  technology  has potential uses in all of these areas. TfI have initially focused their development on low power wireless applications where the benefits of algebraic coding are significant.

 Advantages

This technology offers the following key advantages:

 •  Fast, bespoke code generation.

•  Low encoding and decoding latency.

•  'On the fly' algebraic matrix generation (based on a few initial parameters):

-  Low memory usage.

-  Low power consumption.

-  Low latency

•  Suitable for adaptive coding (dynamically adjusted to the transmission channel). •  Performance comparable to pseudo-randomly generated codes