Hybrid LDPC Decoding With BP Control for Lower Power Reception
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing signal transmission methods using LDPC codes for erasure correction do not effectively manage calculation cost and power consumption in reception devices, leading to inefficient decoding processes.
Innovation Solution
A decoding device is introduced that combines BP decoding and maximum likelihood decoding, with a selector to choose between input signals processed by BP decoding and maximum likelihood decoding, optimizing calculation cost and power consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If BP decoding and Gaussian elimination are combined without controlling decoding in consideration of calculation cost, then decoding accuracy is improved, but power consumption increases
Solution Approach 1:
The patent implements dynamic control of the decoding process by introducing a termination determination unit that monitors decoding progress and decides when to stop BP decoding iterations. This dynamic adjustment allows the system to adapt the decoding effort based on actual performance needs, avoiding unnecessary computational iterations that would increase power consumption while maintaining adequate decoding accuracy.
Solution Approach 2:
The patent changes the parameter of decoding iteration count by introducing a maximum number of iterations limit and a termination determination mechanism. By controlling the number of BP decoding iterations based on termination conditions, the system optimizes the balance between decoding accuracy and computational cost, thereby reducing power consumption without significantly compromising decoding performance.
2Reliability
If more powerful decoding methods are used to improve reception quality, then decoding performance is improved, but calculation cost increases
Solution Approach 1:
The patent applies partial action by using BP decoding with a controlled number of iterations rather than always applying the more computationally intensive Gaussian elimination. The termination determination unit allows the system to perform just enough decoding iterations to achieve acceptable performance, avoiding the excessive computational cost of full Gaussian elimination while maintaining adequate reception quality for many practical scenarios.
Solution Approach 2:
The patent segments the decoding process into two distinct parts: BP decoding as the primary method with controlled iterations, and Gaussian elimination as a fallback for difficult cases. This segmentation allows the system to use the lighter BP decoding for most cases, reducing overall calculation cost, while still having the capability to switch to more powerful Gaussian elimination when necessary to maintain reception quality.
3Reliability
If Gaussian elimination is always applied to ensure decoding success, then decoding reliability is improved, but processing time increases
Solution Approach 1:
The patent applies preliminary action by first attempting BP decoding with a controlled number of iterations before considering Gaussian elimination. This preliminary BP decoding step often suffices to achieve successful decoding, and only when it fails does the system proceed to the more time-consuming Gaussian elimination. This preliminary filtering significantly reduces the average processing time while maintaining high decoding success rates.
Solution Approach 2:
The patent maintains continuity of useful action by implementing a hybrid decoding approach where BP decoding runs continuously with termination conditions, and Gaussian elimination is invoked only when necessary. This continuous monitoring and selective invocation ensures that the system achieves high decoding reliability without the constant overhead of always applying Gaussian elimination, thereby reducing overall processing time.
Data Source
AI summary
A decoding device includes: a BP decoder that performs BP decoding on an input signal; a maximum likelihood decoder that performs maximum likelihood decoding on a signal subjected to the BP decoding; and a selector that selects one of the input signal, the signal subjected to the BP decoding, and a signal subjected to the maximum likelihood decoding. In a configuration of the decoding device, when a decoder is appropriately operated according to quality of data, a calculation scale can be reduced, and power consumption can be decreased.