Next Generation Wireless Networks and Broadband Internet Access

Providing broadband Internet access to all European citizens has been recently acknowledged as one of the main goals ICT should pursue in the near future. In fact, a recent report from EU commission [1] states that, while the broadband use in EU tripled to 36% of households from 2003 to 2007, 7% of the EU�s population is still not connected at all to the Internet, a percentage which becomes 30% in rural areas. Furthermore, there are striking gaps in the EU, where the percentage of population with broadband Internet access ranges from nearly 100% in Denmark, Luxembourg and Belgium, to less than 40% (25% in rural areas) in Romania.

Wireless technologies for broadband data communication are becoming increasingly popular, especially since the advent of cheap hardware based on industrial standards operating in the unlicensed bands such as IEEE 802.11, also known as WiFi. Recently, these technologies have been enriched with the ability of wireless Access Points (APs) to wirelessly communicate with each other (see the drafted 802.11s extension to the standard [2]), forming a wireless mesh infrastructure that can be used by client nodes within the same network to exchange data, and to access the Internet through a few APs providing gateway functionality. This network architecture, known as wireless mesh network (WMN) and shown in Figure 1, has gained significant acceptance in the IT community, and is becoming increasingly popular given its potential to, e.g., share the cost of broadband Internet access in a neighborhood/community, providing ubiquitous Internet access in a city/metropolitan area, bring Internet connectivity to remote rural communities, and so on. Summarizing, WMNs have the potential to become a prominent technology to close the well-known digital divide, and considerably increase the percentage of population having access to broadband Internet, thus fulfilling EU commission goal of providing broadband Internet access to all European citizens.

Figure 1. Wireless mesh network architecture

Next Generation Wireless Networks and MIMO technology

While the density of wireless devices is expected to considerably increase in the next few years, the spectrum available for data communication is expected to remain limited. Hence, the efficient use of such limited spectrum will become increasingly important in the near future. In particular, the challenge to tackle is how to provide ubiquitous Internet access to an increasing number of clients (which leads to a proliferation of wireless devices and surge of traffic demand), while at the same time providing sufficient QoS to each of them, so that the premise of providing broadband Internet access can be fulfilled. The above challenge leads to the need of maximizing traffic carrying capacity of the wireless backbone of a WMN.

In this project, we consider the use of multi-antenna technology to more efficiently use the available spectrum, thus maximizing network traffic carrying capacity. In particular, we envisage using multi-input multi-output (MIMO) systems, whose operating principles are described below. MimoNet goal is to tackle the enormous challenges related to exploiting MIMO technology at a network-level, with the aim at providing a breakthrough increase in network traffic carrying capacity with respect to current wireless technologies. To this end, we will consider the antenna elements available at the wireless nodes as a network resource, rather than simply as a link resource. This shift in perspective entails taking a multi-disciplinary approach to the problem, where network-wide optimization and radio resource management techniques are complemented with the physical layer and signal processing know-how that is necessary to assess feasibility of a proposed approach, and the related technological challenges.

Figure 2. Example of 2X2 MIMO channel

In a MIMO system (Figure 2), the input data stream goes through a preprocessing stage, after which (parts of) the stream is (are) sent to the transmit antenna elements. After going through the wireless channel, which is represented by the so-called MIMO channel matrix H reporting channel gain between all possible pairs of transmit-receive antenna elements, the streams received at the receiver antenna elements are post-processed, so that the original input stream can be correctly recovered. If antenna elements are sufficiently separated, radio signal propagation phenomena such as multi-path fading ensures that the different components of the transmitted signal received at the receive antenna elements can be treated as independent signals, allowing for significant channel capacity (and spectral efficiency) increase. Depending on the specific signal processing techniques implemented, capacity increase can be achieved through either sending multiple concurrent streams between the same transmitter-receiver pair (spatial multiplexing), or nullifying/suppressing interference coming from nearby transmitters so to increase spatial reuse, or combination of them. MIMO technology can also be used to improve channel quality in the low SINR regime through diversity exploitation (beamforming, space-time coding, etc.).

Project goals

To date, MIMO technology has been successfully employed to improve performance of point-to-point communications, e.g., communications between a client and the wireless AP. An example of this is the recent 802.11n extension of the WiFi standard [3], which exploits MIMO links with up to four antenna elements to provide data rates as high as 600Mb/sec on a single link. On the other hand, very little is known to date on how successfully use MIMO technology to provide capacity increase of a whole network, and not a single link. Exploiting MIMO technology at the network-level is an especially challenging problem, which requires considerable, multi-disciplinary efforts from the modeling, algorithmic, analytical, simulative, as well as implementation point of view. For this reason, the MimoNet project proposes undertaking frontier, multi-disciplinary research activities of ground-breaking nature.

Summarizing, the specific MimoNet goals are:

1) designing a suite of protocols for efficient network-level exploitation of MIMO communication technology; the designed protocols will span the MAC, and PHY layer, and account also for network-level issues, with a cross-layer approach;

2) extending the design of state-of-the-art network simulators to enable relatively accurate and reasonably fast performance estimation of medium/large scale (up to a few hundreds of nodes) networks equipped with MIMO technology. The extended network simulator will constitute a fundamental tool to support optimal tuning of the designed protocols under different operational scenarios.

3) carefully estimating protocols performance in selected application scenarios, in a variety of situations for what concerns radio environment (e.g., outdoor urban, free space, indoor, etc.), density of wireless devices (e.g., interference-limited vs. coverage-limited channels), and so on. Performance estimation will be done both analytically (when possible), and through extensive simulation using the extended network simulator developed in the first stages of the project.

4) implementing a representative subset of the designed suite of protocols into a small scale testbed (around 10 mesh routes). The main purpose of the small scale testbed is providing a proof-of-concepts and assess viability of the proposed solutions.

[1] e-Inclusion Newsletter, September 29, 2008.
[3] IEEE 802.11n. Part 11: Wireless LAN Medium Access Control MAC and Physical Layer PHY Specifications: Enhancement for Higher Throughput. Feb. 2007.