Braet and Ballon 2007 - IMS vs E2E
From Inventiopedia
Braet, Olivier and Pieter Ballon (2007) "Strategic design issues of IMS versus end-to-end architectures" in Info, Vol. 9, No. 5, pp 44-56.
This article provides a well-balanced overview over the different arguments for the two competing paradigms of mobile internet architecture: The IP Multimedia System (IMS) versus the end-to-end (E2E) architecture already in place for the PC internet.
The rationale for the IMS is that IP-based hardware has become so cheap that it seems beneficial for mobile operators to exchange the "stovepipe" hardware infrastructure of their existing telecommunication networks, with IP-based hardware. In stead of running IP-based communication on top of voice communication hardware, they would then flip the coin and in stead run voice communication and other services on top of IP-based hardware. This however raises some tecnical challenges, which the IMS is designed to answer. However, the IMS also violates the important end-to-end principles of internet architecture, inserting application-specific functions into the core of the network, allowing network operators to insert gateways and control traffic over the network, favouring certain service over others etc. Not surprisingly the issue is highly controversial, and this article provides a refreshingly non-partisan analysis of the arguments for the two opposing solutions, pointing out strong and weak arguments on both sides.
Finally, the authors propose a series of recommendations for IMS operators, in order for these operators to avoid loosing the competition against E2E alternatives:
- Negotiate new revenue sharing models with content industry
- otherwise, the content industry may prefer P2P distribution models
- Loosen requirements on application developers and device makers
- in order to “ import part of the flexibility and dynamism characteristic of the P2P and internet services environment” - and to avoid being bypassed by VoIP and similar applications - this implies a power shift, towards device manufacturers and content providers - however, application signing and similar guarantee measures seem appropriate
- Share user data with application developers
- giving application developers more control and access to user data, developers can adopt a “fail quick” philosophy, improving speed and innovativeness of application development cycles
- Apply interoperability strategies
- E2E often implies lock-in at application level
- Offer smart identity management
- Competitive advantage: better advertising segmentation
- Support user and service mobility
- primary competitive advantage: TCP/IP does not lend itself to mobile communication
Conclusion:
“As telecommunication companies are moving into the application layer, they will encounter competition from software players that are better acquainted with the faster development cycles of E2E software and the exigencies [fundamental demands] of E2E service deployment, currently enjoy a more rapid market adoption of these services, and can increasingly disregard the telecommunication philosophy of spectrum scarcity.”(54)
--Anders Sundnes Løvlie 10:31, 19 November 2008 (UTC)
