REPORT ON HGWO TO IMPROVE TCP/IP
SOURCES
Point Two [Aim to do this in around about 200-250 words, we have already discussed RoHC in the literature review, as well as VJHC so we can cut down on those bits as much as needed. Perhaps keep them in as background, but only need to be one or two sentences]
Van Jacobson compression
Wildely used, but more effective methods have been created.
Designed for IPv4, and as a switch to IPv6 is happening, this creates some issues unless a new implementation is created
R. Wang and S. Horan “Impact of Van Jacobson header compression on TCP/IP throughput performance over lossy space channels” says that it has severe issues over lossy space environments that decreases performance.
Two other possible methods; Robust Header Compression [RoHC/ROHC], and IP Header compression [D. B. Farouq, A. A. Alarood, N. Aljojo, and A. Abubakar, “Unidirectional and Bidirectional Optimistic Modes IP Header Compression for Real-Time Video Streaming,”]
IP Header Compression [IPHC] boats much more efficient compression than other methods, and has a wider range of implementations than Van Jacobson. Including in this is the fact it is more efficient for TCP/IPv6 [M. Tomoskozi, P. Seeling, P. Ekler, and F. H. P. Fitzek, “Performance Evaluation and Implementation of IP and Robust Header Compression Schemes for TCP and UDP Traffic in the Wireless Context,”]
IPHC does post some concerns with CPU usage, as it is more intensive than Van Jacobson [VJHC] or RoHC
Point Two [Bigger point, perhaps aim for 500 words, depending on how large the first paragraph is]
Making use of multiple cores for TCP/IP
Currently architecture doesn’t make use of multiple core processing for TCP/IP
As Moore’s law is making it so we are switching to including more cores rather than better cores, as making them as such is becoming implausible, making use of these would both make sense as well as make TCP/IP future proof.
Making use of multiple core processing significantly improves throughput [https://ieeexplore-ieee-org.ezproxy.canberra.edu.au/document/6802778] as well as decreasing overall CPU usage
This can dismiss issues with IPHC and its CPU usage
Can have issues on some phone or other portable devices that don’t have multiple cores [rare, but can be an issue.]
Improving efficiency on embedded systems and their alternate processors [ARM, ect. Ect.]
Using the methods and architecture outlined by R. Liao et. al. in their study, we can also increase the efficiency of mobile and other devices.
Solves issue with IPHC and CPU usage
R. Liao, Y. Ji, and H. Li, “Optimized Design and Implementation of TCP/IP Software Architecture Based on Embedded System,”]
Making the switch to a more hardware based TCP/IP infrastructure to improve throughput and decrease power usage could further assist in this process
Using the architecture changes proposed by U. Langenbach et. al [https://ieeexplore-ieee-org.ezproxy.canberra.edu.au/document/6697997] we could further utalizie hardware in order to improve the TCP/IP architecture and it’s overall processing speed.
By improving throughput, and somehow rate limiting output from devices to below the current throughput, this could severely limit congestion as the amount being received is always higher than what is being sent. TCP/IP efficiency improvements can help improve the congestion issue significantly while also minimally effecting overall network speed in respect to uploading.
Last Completed Projects
| topic title | academic level | Writer | delivered |
|---|