Concatenated Truncated Codes Structure Performance and Mechanisation

Anthony R. Pratt and M. Powe

Abstract:	The development of longer code sets for navigation satellite transmissions has resulted in a number of possible alternatives. At the National Technical Meeting in January 2004, a technique using concatenation and truncation of Gold Codes was proposed and first results were demonstrated. The present paper concentrates on the generation of codes of length 10230 chips which could be used for GPS or Galileo services, for example on L5. The codes were chosen on the basis of a number of criteria including the balance, excess line weight, cross- and auto-correlation values, and others. The paper explains the concatenated and truncated code structure which are constructed from Gold codes of length 8k and 2k. For service on L5, a secondary code is used to produce a code of suitable length, being a multiple of 1ms duration with a code chipping rate of 10.23Mcps. The codes are constructed from Gold sequences with certain pre-selected qualities such as being nearly balanced. The paper shows how the preferred pairs of m-sequence polynomials are selected and shows the various imbalance characteristics of these sequences. The work reported here establishes the scope of choice for optimisation of the codes including code matching and starting point selection. The complexity of the search process for good codes is quantified. One of the selection criteria attempts to control the effects of near multi-path propagation through choosing codes with low values of auto-correlation coefficients for delays within a few chips of the correct replica alignment. This has a limiting effect on the number of codes available. Probabilistic models are used to establish the numbers of remaining code families. Formulae are derived showing how many codes and code families exist within the constraints of 8k + 2k code construction. The paper sets out details of the code construction methods and gives examples of such codes. The performance attributes used for code selection are set out in detail and the performance of the actual codes against these criteria are given. These include auto- and cross-correlation histograms at various Doppler offsets, excess line weight and values for the low delay auto-correlation functions. The performance of the codes shown may be considered typical for the class. Several possible mechanisations for code generation are disclosed which might be employed by receiver manufacturers, at least as starting point for the design process. The paper concludes with some further possibilities for code development within the scope of a generalised Gold structure.
Published in:	Proceedings of the 2005 National Technical Meeting of The Institute of Navigation January 24 - 26, 2005 The Catamaran Resort Hotel San Diego, CA
Pages:	741 - 753
Cite this article:	Pratt, Anthony R., Powe, M., "Concatenated Truncated Codes Structure Performance and Mechanisation," Proceedings of the 2005 National Technical Meeting of The Institute of Navigation, San Diego, CA, January 2005, pp. 741-753.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In