A huge amount of generated data is regularly exploding into the network by the users through smartphones, laptops, tablets, self-configured Internet-of-things (IoT) devices, and machine-to-machine (M2M) communication. In such a situation, satisfying critical quality-of-service (QoS) requirements (e.g., throughput, latency, bandwidth, and reliability) is a large challenge as a vast amount of data travels into the network. Nowadays, strict QoS requirements must be satisfied efficiently in many networked multimedia applications when intelligent multi-homed devices are used. Such devices support the concept of multi-homing. To be precise, they have multiple network interfaces that aim to connect and communicate concurrently with different networking technologies. Therefore, many multipath transport protocols are provided to multi-homed devices, which aim (1) to take advantage of several network paths at the transport layer (Layer-4) and (2) to meet the strict QoS requirements for providing low network latency, higher data rates, and increased reliability. To this end, this survey first presents the challenges/problems for supporting multipath transmission with possible solutions. Then, it reviews recent research efforts related to the concurrent multipath transmission (CMT) protocol and the multipath transmission control protocol (MPTCP). It reviews the latest research efforts by considering (1) how a multipath transport protocol operates (i.e., its functionality); (2) in what type of network; (3) what path characteristics it should consider; and (4) how it addresses various design challenges. Furthermore, it presents some lessons learned and discusses open research issues in multipath transport protocols.

The Internet of Things (IoT) is going to create a heterogeneous network of billions or trillions of connected things with self-aware, self-configuration capabilities that will exchange data with each other. This heterogeneous IoT network requires standard protocols, secure and energy-efficient wireless communication, and the allocation of unique identities such as IPv6 addresses. One of the key challenges in such a heterogeneous network is the maintenance of device location and unique identity to ensure the smooth functioning of location-aware applications. Hence, this paper presents a stateless spatiotemporal IPv6 addressing scheme called four-dimensional IP addressing (FDIPA). It is founded on the idea that two objects can not be in the same place at the same time. It uses the time and location information of the device to generate the target IPv6 address and assigns it to the device after verifying its uniqueness with the duplicate address detection (DAD) protocol. It not only maintains spatial and temporal information but also guarantees a unique identity for a large-sized network. Analytically, it is proved that FDIPA will always generate a unique identity if there is no technological limitation. However, the performance of FDIPA is evaluated based on metrics such as Assignment Success Rate (ASR), communication overhead, and total energy consumption in the assignment of a unique IPv6 address. The experimental results show that FDIPA achieves good ASR while keeping low energy consumption and less overhead in the network.

In the IoT network, each node should have the self-configuration capability as well as a unique identifier such as IPv6. The unique IPv6 identity of nodes is primarily maintained by employing the stateless auto-configuration method and Duplicate Address Detection (DAD) protocol. In the existing DAD solutions, when a new node creates a target IPv6 address, it broadcasts a full or partial target address across the network to verify uniqueness. In this process, when the entire target address is broadcast, the target address is exposed to all the nodes of the network, whereas in partial broadcasts, the address of some nodes of the network is exposed to the new node. A malicious node can take advantage of this feature and can disrupt the DAD process by claiming the target address or by making the existing address as assigned. To protect the DAD process from such an attack, this research paper presents a better and secure DAD process, in which, a new node and the existing node transmit only a partial address and both nodes preserve a portion of the address. In this way, neither the existing nor the new malicious node can know the full address of each other, so they can not disrupt the DAD process. Experimental results indicate that the address success rate (ASR) of the proposed scheme outperformed existing DAD solutions in the presence of a malicious node. Moreover, the only partial address is transmitted by the nodes in the network hence reduces addressing cost and energy consumption.

In today's world mobile agents and mobile adhoc networks are the two technologies that are contributing towards better connectivity and communicability. When the two technologies used jointly, the interest is increased. Due to the properties like selfconfiguration and infrastructure-less, the mobile ad-hoc networks provide various remarkable features. But the various challenges are also associated with mobile ad-hoc networks like dynamic topology, mobility, energy constraint etc. Mobile agent provides solution to these challenges. A mobile agent is a new way of computer interactions and provides better options for the developers to create applications based on connectivity. Mobile agents move around the ad-hoc networks in a different and better way than the other widespread client server architecture based applications. Due to their mobility and autonomy these agents can perform various functions in mobile ad-hoc networks like topology discovery, routing, key management, congestion control etc. The paper reviews the role of the mobile agents in the mobile ad-hoc networks, and emphasizes its application on the various layers of a layered architecture of mobile ad-hoc networks and concludes its merits as compared to other conventional approaches.