Erlang C for landline telephone traffic engineering

• Main Author: Aurelio, Aldwin D.
• Format: text
• Language: English
• Published: Animo Repository 1999
• Subjects: Queuing theory; Telephone; Telecommunication > Traffic; Electrical and Electronics
• Online Access: https://animorepository.dlsu.edu.ph/etd_masteral/2029
• Institution: De La Salle University

This study seeks to determine the application and the viability of Erlang C traffic table as compared to presently used Erlang-B in determining the necessary telephone circuits in landline telephone traffic engineering.Erlang B and Erlang C were applied on a fully symmetric, non-hierarchical, circuit-switched network utilizing a fully digital, computer-based switch. The performance measures used were end-to-end blocking probability and average blocking probability. These performance measures analyzed the response of Erlang B and Erlang C based on network variables such as frequency of call retries, overload in the system, number of circuits used, and number of alternate routes.Erlang C requires more circuits than Erlang B for the same grade of service. For a given grade of service and at low traffic load, a minimal difference exists between circuits required for Erlang B and Erlang C that is usually one to two. But at higher traffic load it becomes more pronounced, especially if the blocking probability is higher than 1 percent (say 5 percent). In Erlang-B-assumed network, Erlang C sees a higher blocking than what is registered by Erlang B. Conversely, in Erlang C networks, Erlang B sees a lower blocking instead. This means that while Erlang B is saying a relatively constant blocking percentage over an entire range of traffic load, Erlang C says it is much higher than that. Erlang C specifically takes in the probability that calls on encountering a blocked circuit makes a reattempt until it gets through or is saved. This means that any entry on the part of the caller will not contribute to the increase in fresh traffic offered to the network because retried calls go to a virtual queue by virtue of a continuous redial and remains in there until a free path becomes available. However, Erlang B sees it differently, because it assumes that calls on encountering a blocked circuit do not make a reattempt and are lost. Any reattempt on the part of the caller is seen as a new call and thus contributing to the increase in offered traffic. Thus Erlang C can serve more subscribers while maintaining within the limits of the grade of service.