Ricerc@Sapienza

The C-MIANS platform requires minimum modifications to the ANDSF and MIH protocols, and a minimum overhead of control messages exchange, due to the introduction of the IEEE 1900.6 sensing architecture. Moreover, C-MIANS does not increase the energy consumption of the communication network, because the sensing activities are performed in a dedicated architecture. The C-MIANS procedure can be either network or user-driven, and it is triggered by the events defined within the MIH event service.
The procedure foresees five steps:
* Step 1 (ANDSF activation): the user connects to the ANDSF server, and receives operator policies and coarse context information.
* Step 2 (MIH and IEEE 1900.6 spectrum sensing activation): with a single message, the user triggers fine context retrieval to its PoS, and requires its nearby spectrum agents to perform spectrum sensing on segments occupied by the candidate networks.
* Step 3 (Context retrieval and candidate networks evaluation): the PoS receives fine context information about candidate networks (MIH), and sensing results (IEEE 1900.6), deactivating spectrum agents once all spectrum segments have been analyzed. By combining sensing and context information, the PoS evaluates optimal access configurations for each candidate network, and transfers this information to the user, acting as data archive and cognitive engine from an IEEE 1900.6 perspective.
* Step 4 (Network selection): the user selects the best network by combining the access configurations, provided by the PoS, with the ANDSF operator policies, and informs the ANDSF server about the decision, enabling a fast update of the operator policies. Note that feedback towards the ANDSF server is not part of the original ANDSF protocol, but might prove extremely useful in specific situations that require real-time policies update, such as crowded areas and high mobility communication scenarios.
The large scale deployment and the adoption of additional spectrum segments for smallcells, including unlicensed spectrum bands, as for example ISM bands at 2.4 and 5 GHz, typically used by WiFi, have been proposed in 4G [6]. Tentatives of exploiting the WiFi unlicensed spectrum, mainly known as LTE-unlicensed (LTE-U) or licensed-assisted access using LTE (LAA-LTE), generate the problem of internetwork interference. A dynamic time-domain resource sharing has been proposed for LTE and WiFi coexistence. In this approach, the LTE-U network istantiates time frames with reduced data and power; the WiFi APs apply then spectrum sensing to detect those frames, and use them for their own transmissions. The scheme, referred to as almost blank subframes (ABSs) [6], can be improved by optimizing sensing thresholds and duration, as well as by enabling LTE to estimate activities and densities of nearby WiFi APs [7].
In 5G, the challenge posed by coexistence of HetNets and WiFi will expand beyond the ISM band, to the 60 GHz mm-Wave bands, where the new IEEE 802.11ad standard, also known as WiGig, is expected to coexist with 5G backhaul links between SBSs and the core network, required to efficiently support the dense deployment of smallcells. In the 5G vision, interoperability schemes between cellular and WiFi systems must therefore be supported by internetwork interference mitigation schemes.
The proposed C-MIANS platform has the potential for solving the above challenge while merging interoperability and interference mitigation.

[6] H. Zhang et al., "Coexistence of Wi-Fi and Heterogeneous Small Cell Networks Sharing Unlicensed Spectrum," IEEE Commun. Mag., vol. 53, no. 3, 2015, pp. 158-164.
[7] G. Guanding et al. (eds), "LTE in Unlicensed Spectrum," IEEE Wireless Commun., Special issue, vol. 23, no. 6, 2016.