Host mobility has traditionally been worked out at the network level, but even for the long years of standardizing Mobile IP, this essential thing has not been supported by operators. The double function of IP as a location detector and communication endpoint identifier provides a number of functional and performance troubles.
Due to these scenarios, IT specialists agreed that the best place to manage the mobility problem is at the transport layer. Though it is not a fresh idea, they believe that the surfacing standard of MultiPath Transmission Control Protocol (MTPCP) can be implemented to solve various issues connected to mobility. MPTCP naturally applies make-before-break, can be incrementally installed, is backwards compatible with standard TCP, and could even simplify incremental adoption of IP v 6.
The Call for MPTCP Adoption
It has become a usual thing for mobile devices like tablets and smart phones to have multiple radios such as 3G, WiFi, and Bluetooth. With developing integration and software-defined radio, we anticipate this trend to go on, and for devices to prop up more radio technologies. Each radio technology has its benefits and drawbacks; 3G brings more omnipresent coverage, but in big cities often experiences congestion to be almost impractical; WiFi gives high transfer speeds, but coverage is inconsistent and for mobile users is often transient. So what shall we do to utilize the available links in order to acquire the best coverage and throughput, and use the slightest battery power while doing it?
Conventional remedies are basic; smartphones normally connect via a single network at a time using an established policy such as “use WiFi if available, otherwise 3G”. While WiFi and 3G networks each provide an IP address, transition between the two is unsettling, requiring applications to re-launch connectivity. Such rudimentary policies work reasonably well if the user is not mobile. When in a car or train, WiFi connectivity is transient – usually good connectivity is accessible but only for a while. Such procedures are too disruptive to utilize when connectivity (or loss of connectivity) is transient.
How Does MPTCP Works?
The concept behind multipath TCP is to use additional paths that are disregarded by the routing system. By doing such procedure, it can provide more bandwidth and better resiliency for the user. It also gives superior network utilization for network operators.
MTPCP involves adjusting TCP to give it the capacity to send a specified packet over a given path. Then, it operates the regular NewReno congestion control algorithms discretely for each path, so each path has its own congestion window that reveals the available bandwidth of the path. As a result, MPTCP can transmit many packets over paths with a small RTT or round-trip-time and a large congestion window. It also sends fewer packets over paths with a large RTT and/or small window. This process enables the multipath TCP to adjust instantly and quickly to network congestion, moving away the traffic from crowded paths and toward uncongested paths. Routers can also share out packets over several paths, however routers doesn’t possess end-to-end congestion data. Further, routing protocols that alter routing based on link usage have never earned a grip since they can cause detrimental oscillations as they convey traffic to idle links, which can then become congested, move traffic back, and so on.
The Benefits of MTPCP to Different End-user Groups
This foremost advantage is most applicable to heavy users which mean it is very useful for domestic use, sites and human utility. Indirectly, it is also a beneficial thing to providers as it gives an insight of a better service. For content providers, superior throughput also potentially means more sales. It is essential to remember that these advantages do not openly generate revenue for end-users. They are instead considered as enhancements that bring in more business for suppliers, more productive time for companies and less hassle for domestic users through greater availability or better perceived service.
This positive thing about MPTCP is primarily of interest to both heavy and light mobile users. This innovation allows the session to continue in the event there are deliberate changes in connectivity.
The adaptability and flexibility of MPTCP is relevant to all kinds of clients/consumers. However, larger sites will practically see a greater benefit since there are more people that might be inconvenient in the case of failure. In addition, the need of companies for business stability increases their demand for resilience.
MPTCP and Apple’s Next Gen Operating System
Apple’s inclusion of Multipath TCP in iOS 7 is such a big buzz and thought to be the first consumer product to integrate with the technology. As its name clearly means, MPTCP allows for an iPhone or iPad or any connected device to transmit data over multiple pathways at the same time. For instance, this adaptation enables for compatible gadgets to transport data over Wi-Fi and 3G networks. This gives the ability to the protocol to push data throughout the most competent network, which leads to fewer failures. So if a single channel fails, another channel will take over. This brilliant discovery was pioneered by Olivier Bonaventure, a computer science professor working at the IP Networking Lab in Belgium.
MultiPath TCP took its origin in 2008 when European Commission funded and initiated a so-called Trilogy Project. However, this technology has yet to see a pervasive adoption and for sure, if utilized by a majority of end-users, will change the future of mobility and connectivity.