Fabini, J., Hirschbichler, M., Kuthan, J., & Wiedermann, W. (2013). Mobile SIP: An Empirical Study on SIP Retransmission Timers in HSPA 3G Networks. In T. Bauschert (Ed.), Advances in Communication Networking 19th EUNICE/IFIP EG WG 6.6 International Workshop, Chemnitz, Germany, August 28-30, 2013, Proceedings (pp. 78–89). Springer LNCS/Springer. https://doi.org/10.1007/978-3-642-40552-5_8
-
Book Title:
Advances in Communication Networking 19th EUNICE/IFIP EG WG 6.6 International Workshop, Chemnitz, Germany, August 28-30, 2013, Proceedings
-
Abstract:
Mobile packet-switched voice must replace mobile circuit-switched voice latest with the large-scale deployment of LTE. However, short-term full-area LTE coverage is highly unlikely, handover to and interworking with 3G HSPA networks being a must. This raises the question of how the Session Initiation Protocol SIP performs in today's HSPA networks.
Relying on live SIP provider monitoring data and active HSPA measurements, this paper demonstrates that default SIP retransmission timers are inappropriate for real-world deployments of packet switched voice over today's 2G and 3G networks including HSPA. Main reason for this deficiency is the reactive, demand-driven resource allocation strategy of HSPA which results in huge uplink delays. These delays eventually lead to a high number of unnecessary SIP message retransmissions which load networks and servers. SIP clients for mobile devices should therefore implement adaptive, user-configurable SIP retransmission timers and provide appropriate default values for these timers to enable seamless SIP operation in mobile cellular networks.
Mobile packet-switched voice must replace mobile circuit-switched voice latest with the large-scale deployment of LTE. However, short-term full-area LTE coverage is highly unlikely, handover to and interworking with 3G HSPA networks being a must. This raises the question of how the Session Initiation Protocol SIP performs in today's HSPA networks.
Relying on live SIP provider monitoring data and active HSPA measurements, this paper demonstrates that default SIP retransmission timers are inappropriate for real-world deployments of packet switched voice over today's 2G and 3G networks including HSPA. Main reason for this deficiency is the reactive, demand-driven resource allocation strategy of HSPA which results in huge uplink delays. These delays eventually lead to a high number of unnecessary SIP message retransmissions which load networks and servers. SIP clients for mobile devices should therefore implement adaptive, user-configurable SIP retransmission timers and provide appropriate default values for these timers to enable seamless SIP operation in mobile cellular networks.
