Fig 1.
Triangle routing in mobile IP.
Fig 2.
Standard TCP and MPTCP protocol stacks.
Fig 3.
Initiating an MPTCP connection.
Fig 4.
Starting new subflow.
Fig 5.
(a) Address addition, (b) Address removal.
Fig 6.
Transfer of data using MPTCP.
Fig 7.
Change in MPTCP subflow’s priority.
Fig 8.
The closing of the MPTCP connection.
Fig 9.
Hidden Markov model.
Fig 10.
Enhanced handover mechanism using mobility prediction.
Fig 11.
Mobility prediction stages.
Fig 12.
First HMM represented by location.
Fig 13.
Second HMM represented by location and WiFi AP.
Fig 14.
Visualization of WiFi signal strength.
Fig 15.
Proposed progressive mobility prediction.
Fig 16.
Example of the dual hidden Markov model.
Fig 17.
Homogeneous handover between WiFi APs.
Fig 18.
Seamless heterogeneous handover.
Fig 19.
Use of the cellular network as the primary network path.
Fig 20.
Use of the WiFi network as the primary network path.
Fig 21.
Comparison of network throughput between eHMP and EOPAPS.
Table 1.
Average throughput between eHMP and EOPAPS in each range.
Fig 22.
Instantaneous RSSI of EOPAPS and the proposed eHMP.
Fig 23.
The retransmission rate using eHMP and EOPAPS.
Fig 24.
MPTCP capable message exchange between the mobile device and server.
Fig 25.
New subflow for data communication.
Table 2.
Average throughput of MPTCP and MPTCP with mobility prediction in each range.
Fig 26.
The network throughput of MPTCP and MPTCP with mobility prediction.
Fig 27.
RSSI of connected WiFi AP in MPTCP and MPTCP with mobility prediction.
Fig 28.
The average retransmission rate using MPTCP and MPTCP with mobility prediction.
Table 3.
Average throughput of MPTCP and heterogeneous handover using eHMP in each range.
Fig 29.
Network throughput of MPTCP and heterogeneous handover using eHMP.
Fig 30.
RSSI of connected WiFi AP in MPTCP and heterogeneous handover using eHMP.
Fig 31.
The average retransmission rate using the proposed eHMP, MPTCP and MPTCP with mobility prediction.