Spectrum is life-blood of wireless networks. Licensed spectrum is operators’ first choice, as it provides reliability, and predictable performance. Hence the industry is hard at work to make best use of existing spectrum and looking for more licensed spectrum via traditional auctions and innovative paradigms such as authorized/licensed shared access (ASA/LSA). At the same time, to support the insatiable data demand, operators will also have to leverage readily available unlicensed spectrum. Bringing LTE Advanced to unlicensed spectrum is the most efficient option to achieve that.

Bringing LTE Advanced to unlicensed spectrum is really a simple idea with immense benefits. It involves leveraging the large number of small cells that operators are planning to deploy and aggregating unlicensed spectrum with the licensed spectrum for LTE Advanced. The existing core network can be used as is. In essence, the whole system works as a unified LTE network to efficiently leverage both licensed and unlicensed spectrum bands.

LTE Advanced in the unlicensed bands can provide better coverage and capacity than Wi-Fi deployed by network operators (also referred to as Carrier Wi-Fi), while allowing for seamless flow of data between licensed and unlicensed spectrum through a single core network. This means operators can augment the capacity of their networks by utilizing the unlicensed spectrum more efficiently, while also providing the tightest possible interworking between the licensed and unlicensed bands. From the user perspective, this means an enhanced broadband experience—higher data rates, seamless use of both licensed and unlicensed bands, higher reliability, better mobility, and more. LTE Advanced in unlicensed bands is carefully designed to harmoniously co-exist with Wi-Fi. There are many features to avoid and mitigate interference, as well as to share the resources proportionately and fairly, when using the same channel, all-in-all to be a “good-neighbor” to Wi-Fi.

Source : Qualcomm

MPTCP is an evolution of TCP that allows a concurrent use of multiple interfaces for a single TCP session. If the both sides of a TCP connection are MPTCP-capable (otherwise a legacy TCP will be used between them), multiple subflows are established over different data paths, i.e., multiple network interfaces, and the MPTCP sender spreads TCP segments over the subflows. Spreading data over subflows is enabled by defining and managing separate congestion windows for individual subflows. TCP’s congestion control mechanism then distributes traffic based on the congestion condition of each data path. When a high congestion condition is detected in a data path, the amount of traffic on it is reduced and other less congested paths convey more. Thus MPTCP accomplishes dynamic load balancing between interfaces.

The detail explanation about MPTCP can be read below.