Comparison of Two Approaches for GNSS Receiver Algorithms: Batch Processing and Sequential Processing Considerations

Receiver design implementations are considered in terms of batch vs. sequential processing strategies. It is shown that the batch processing techniques improve performance characteristics of GNSS signal tracking for a number of application areas. Particularly, flight test results presented demonstrate that batch processing improves the GPS tracking margin by 8 dB as compared to sequential tracking for the deep GPS/INS integration mode that performs code and carrier phase tracking, and data bit recovery. INTRODUCTION Since the processing of digitized Global Navigation Satellite System (GNSS) signals has become feasible the number of terms used to describe different processing steps and different processing strategies has been increasing. Example terms include software-defined radio [1], software receiver [2], frequency-domain tracking [3], tracking loops [4], block processing [5], parallel FFTbased code search [6], parallel FFT-based frequency search [7], high sensitivity receivers [8], ultra-tight integration [9], deep integration [10], massive parallel correlator banks [11]. As the number of processing terms increases, their place in a framework for the GNSS receiver is becoming somewhat obscure. Hence, the first goal of this paper is to show the place of existing GNSS signal processing terms in the receiver design flow. This goal is achieved by representing the receiver design using sequential and batch processing approaches since the existing signal processing terms fit into sequential, batch or a combination of sequential and batch processing strategies. It is important to mention that sequential processing is well recognized while batch processing techniques remain underutilized despite the fact that for a number of practical applications the use of batch processing techniques efficiently overcomes limitations of the sequential processing approach. The second goal of this paper is therefore to demonstrate application areas for which batch processing is more powerful than a sequential approach. Example applications include tracking of signals with a low carrier-to-noise ratio (CNR) required to perform indoor navigation, navigation under foliage, and navigation in interference environments. The deep Global Positioning System (GPS)/Inertial Navigation System (INS) integration is used as a case study for these application examples. Goals of this paper are thus formulated as follows: Show the place of existing processing terms in a framework for the GNSS receiver design. Provide new material that demonstrates efficient use of batch processing techniques for practical application areas. The remainder of the paper is organized as follows. First, a simple filtering example is used to introduce basic concepts of sequential and batch processing. Basic sequential and batch processing concepts are then applied to consider generalized sequential and batch architectures of GNSS signal processing. The main result of this consideration is that the sequential and batch processing approaches have complementary features. Receiver design implementations that utilize combinations of sequential and batch processing techniques are therefore presented next. Finally, deep GPS/INS integration is used as a case study to demonstrate advantages of batch processing over sequential implementations. INTRODUCTION TO SEQUENTIAL AND BATCH PROCESSING: FILTERING EXAMPLE This section illustrates basic concepts of sequential and batch processing approaches using a filtering example. The example is formulated as follows: smooth 5 20 ION GNSS 18th International Technical Meeting of the Satellite Division, 13-16 September 2005, Long Beach, CA measurements given the measurement model defined by Equation (1): 5 ,..., 1 k , n a x ~ k k = + = (1) where: 5 ,..., 1 k , x k = is the measurement; const a = ; 5 ,..., 1 k , nk = is the measurement noise. Figure 1 illustrates the estimation process of a sequential filter for the filtering example considered.