Burst Error Detection Using Sliding Window Comparators
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current network systems fail to detect all burst error conditions in packet networks due to the arbitrary selection of check windows, leading to missed error reports and impractical hardware implementation.
Innovation Solution
A burst error detection device measures the interval between incoming packets, stores these values in a memory chain, and uses comparators to determine if a burst error condition has occurred by comparing the accumulated values to pre-configured periods, generating an alarm for both under and over-received packets.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If the slipping window method is used to monitor packet jitter, then the implementation is simple and attractive for hardware, but burst errors are not detected on every possible check window leading to missed error reports
Solution Approach 1:
The patent divides the monitoring function into multiple parallel comparators, each responsible for a specific check window position. Instead of using a single slipping window that checks only predetermined positions, the system segments the detection task across multiple comparators (e.g., COMP1, COMP2, COMP3) that simultaneously monitor different window positions. This segmentation ensures that burst errors are detected regardless of their position in the packet stream, while maintaining hardware feasibility through parallel simple comparison operations.
2Reliability
If sub-windows are made infinitely small to detect all possible burst errors using the slipping window method, then detection completeness improves, but the method becomes impractical to implement
Solution Approach 1:
The patent introduces dynamic positioning of check windows through multiple comparators that can be selectively activated. Instead of using a single static slipping window or infinitely small sub-windows, the system dynamically configures multiple comparators to monitor different window positions simultaneously. This dynamic approach achieves complete burst error detection by adapting the monitoring positions to cover all possible error locations, while maintaining practical implementation through a finite number of configurable comparators rather than requiring infinitely small divisions.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Presented is an apparatus (200, 300, 400) and methods (500) for detecting any burst error condition on a network (204). The apparatus (200, 300, 400) measures the interval (206, 306, 406) between arriving packets (304, 404) on a network (204) and stores (208) a series of interval measurements for analysis. The apparatus (200, 300, 400) also provides a comparator (212, 322, 324, 426, 428) for determining if a burst error condition (330, 332, 434, 436) occurred. The method (500) shifts the stored interval values (306, 308, 310, 312, 314, 316, 406, 408, 410, 412, 414, 416) based on the arrival (304, 404) of the next packet and its associated interval value and sums (320, 418, 420, 422) a subset of the stored interval values (306, 308, 310, 312, 406, 408, 410, 412) for comparison to a pre-configured period (328, 432) to determine if a burst error condition (332, 436) exists for too few packets. The method (500) provides the sum (318, 418, 420, 422, 424) of a separate subset of stored interval values (308, 310, 312, 314, 316, 406, 408, 410, 412, 414, 416) to another comparator (322, 426) for comparison to a pre-configured period (326, 430) to determine if a burst error condition (330, 434) exists for too many packets.