Bit Error Rate Estimation for Sparse and Burst Errors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for estimating bit error rates (BER) in network devices are unreliable, especially when bit errors are sparse, leading to inaccurate results and inefficient use of processor resources, particularly during error bursts.
Innovation Solution
The implementation determines an estimated BER by using a division model when errors are frequent and a statistical approach when errors are sparse, adjusting for error bursts by comparing observed bit error counts with expected counts across different window lengths, and conserving resources by treating error bursts as single errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a division model is used to estimate BER by dividing bit error count by quantity of bits, then the estimation process is simple, but the accuracy deteriorates when bit errors are sparse
Solution Approach 1:
The patent applies dynamics by making the BER estimation method adaptive rather than static. The system dynamically selects between different estimation approaches (division model vs. statistical approach) based on the observed bit error rate characteristics. When errors are frequent, it uses the simple division model; when errors are sparse, it switches to the more accurate statistical approach, allowing the estimation method to adapt to changing error conditions
Solution Approach 2:
The patent changes the estimation parameter from a single division-based calculation to a multi-parameter statistical model that incorporates error counts, bit quantities, and probability distributions. By introducing these additional parameters and using hypothesis testing, the system achieves higher accuracy in sparse error conditions while maintaining computational feasibility
2Measurement precision
If burst check bit error counts are obtained after detecting an error burst, then the accuracy of BER estimation improves, but the processing time increases
Solution Approach 1:
The patent applies preliminary action by pre-establishing statistical models and thresholds for error burst detection. When an error burst is detected, the system has pre-configured burst check intervals and statistical criteria ready, allowing for rapid assessment without extensive real-time computation. This preparation in advance reduces the processing time required when actual bursts occur
Solution Approach 2:
The system uses periodic burst check intervals to monitor error patterns without continuous analysis. By sampling at predetermined intervals and comparing against statistical thresholds, the system achieves accurate BER estimation during error bursts while avoiding the need for continuous processing, thus reducing overall processing time
3Measurement precision
If a statistical approach with sliding windows is used to estimate BER, then the accuracy improves for sparse errors, but the device complexity increases
Solution Approach 1:
The patent applies segmentation by dividing the bit stream into sliding windows of fixed size and analyzing error patterns within each window independently. This segmentation allows the complex statistical analysis to be performed on manageable portions of data rather than the entire stream, reducing computational complexity while maintaining accuracy. Each window can be processed separately with predefined statistical parameters
Solution Approach 2:
The system uses disposable sliding windows that are created, processed, and discarded sequentially rather than maintaining complex persistent data structures. Each window is a temporary, lightweight object that requires minimal memory and processing overhead, enabling accurate statistical analysis without significant device complexity
Data Source
AI summary
A system may obtain a current bit error count that identifies a quantity of bit errors in a bit stream during a time interval. The system may determine that the current bit error count identifies one or more bit errors. The system may determine whether an estimated bit error rate (BER) for the bit stream is likely to satisfy a threshold. The system may select an approach for determining the estimated BER for the bit stream. The estimated BER may be determined based on combining the current bit error count with a quantity of bits received in the time interval when the estimated BER is likely to exceed the threshold, and the estimated BER may be determined based on the current bit error count and one or more past bit error counts when the estimated BER is unlikely to exceed the threshold. The system may determine the estimated BER.