After all the trouble I had with ns-2, I’ve gotten round to reading my first paper. It was an interesting read, but I felt it was a bit thin on content. I’ve gotten used to the idea that academic papers tend to be on the dry side, but all the advice I’ve had thus far is to persevere. So it was with a certain relief that I found a paper that was easy to read! A summary of what the contents along with a brief critique is below. One down, eight to go!
Clincy, V. A. and Mudiraj, P. 2006. The future leading mobility protocol: mobile IPv4 OR Mobile IPv6?. J. Comput. Small Coll. 22, 1 (Oct. 2006), 197-203.
The paper looked to evaluate both MIPv4 and MIPv6 so as to give an opinion as to which protocol would be the ‘Future Leading’. What the article failed to comment on was the fact that the world is fast running out of IPv4 addresses which may have an impact on its suitability as a ‘Future Leading’ protocol.
The paper gave a brief overview of the key differences between the two protocols:
- MIPv4 requires the use of a Foreign Agent
- MIPv6 uses Route Optimization (sic) for improved performance
- MIPv4 uses ARP
- MIPv6 uses Neighbour Discovery
- MIPv4 uses a directed broadcast approach to discover the Home Agent
- MIPv6 uses anycast and receives a single reply
The paper mentioned that one of the key aspects of performance is in the handover from one access point to the next whilst a mobile node is moving. The details of this differ from MIPv4 to MIPv6. It is worth mentioning that IPv6 has had Mobility in mind since inception.
My chief criticism with this paper is that the comparison experiments were limited to two scenarios but neither used both protocols for a fair test. As mentioned before, both IPv4 and IPv6 implement Mobile IP differently. The scenarios were:
- Wired Mobile IP network using Mobile IPv4
- Wireless Mobile IP network using Mobile IPv6
Surely it would have made sense to at least run MIPv4 and MIPv6 across a wired and wireless Mobile IP network? Also the chosen metric to assess the performance was the end-to-end delays. No explanation was given which would have been insightful.
Finally there was some data supplied which suggested that MIPv6 is more efficient at routing information as a result of there being no foreign agent and the Route Optimisation strategy.
So in summary, I didn’t learn much more than I’d already found out from books and in the end, the paper didn’t commit to what the authors felt was a ‘Future Leading’ protocol.