Performance improvements of the IEEE 802.2 LLC type 2 protocol

The author investigates the effects on performance of three different modifications to the IEEE 802.2 LCC type 2 protocol (further referred to as ORIG) in an internetwork environment consisting of three IEEE 802.5 token-ring networks connected via bridges to a backbone ring. The modifications are: (1) selective reject of lost or erroneous frames; (2) acknowledgement accumulation with a static threshold value; and (3) acknowledgement accumulation with a dynamic threshold value. The prime performance measures are throughput and mean end-to-end delay. It is found that modifications (1) and (3) yield a significant improvement in performance under a high traffic load as compared to ORIG and that modification (2) increases the throughput at the cost of a higher mean end-to-end delay