Wireless Mesh Technology and Network

This work was funded by the Natural Science Foundation of China under Grant 60572066.

     The Wireless Mesh Network (WMN) is a new wireless mobile communications technology for access to IP network anywhere. It is applicable to regional environmental coverage and broadband high-speed wireless access. It is based on mutual cooperation and coordination among a variety of wireless access points in mesh distribution, which boasts the advantages of broadband high-speed and high spectrum efficiency, and features dynamic self-organization, self-configuration and self-maintenance. Therefore, the research, development and application of wireless Mesh technology and network has become one of the hottest topics on  wireless mobile communications, especially during the standardization research on IEEE 802 series and in the Long Term Evolution (LTE) of future mobile communications systems, wireless Mesh technology and network have become an eye-catching focus[1-6].

1 Wireless Mesh Technologies
Besides a general Wireless Gateway (WGW, which implements wireless Internet access) and wireless user terminals, a WMN has an additional wireless router so that the original wireless access network architecture, as shown in Figure 1, evolutes into the WMN architecture, as shown in Figure 2. A wireless router layer is added to the WMN. Wireless connection is made between different routers and between routers and wireless IP access points (WGW), and furthermore, cross connection is supported. Thus, a dense network is formed in this way. Hence, the unique basic technologies and processing methods of WMN are derived. All of them are directly associated with the new wireless router layer, which makes them different from the conventional pure wireless access networks.

1.1 Wireless Transmission Technology of  Wireless Mesh Router
During the research of WMN technologies, the wireless transmission technology of Mesh router, such as Wireless Router (WR), is normally referred to as the “physical layer technology of WMN”. Here, the term “transmission” refers to wireless transmission between WR and user terminals, between WRs, and between WR and WGW.

     The wireless transmission between WR and user terminals implements functions similar to base stations or wireless access points and can meet the access requirement of diversified wireless air interfaces. WMN architecture supports different standard access systems, and different wireless transmission technologies to fulfill the wireless transmission between WR and user terminal.

    The wireless transmission between WRs and that between WR and WGW need to be defined and confirmed. In principle, there is no direct relationship between the selection of transmission technologies and the technical standards and system modes supported by user terminals; however, it’s better to use the existing state-of-the-art technologies and methods. A WR is equivalent to a base station. It has fixed location, supports integrated high-speed data of multiple users, features dense coverage, needs prevention of mutual interference as much as possible, and provides selection of multiple routes, so smart directional antenna technology, efficient controllable modulation and coding technology, and low critical transmit power control technology are the most important physical layer transmission technologies.

    The smart directional antenna technology is a beamforming technology in a designated direction with concentrated signal power, as shown in Figure 3. This technology has found wide application in 3G systems, especially in Time Division Synchronous Code Division Multiple Access(TD-SCDMA) systems in the operating mode of shared frequency. The smart antenna technology is a special Multiple-Input Multiple-Output (MIMO) technology which uses the combination of m phase-controlled antenna dipoles to generate low-power directional transmission in m  different directions so that the signal arriving at the access point has the strongest power but exerts minimum radiation and influence on other adjacent WRs, thus, to implement
low-power application of dense network coverage. In applications where a smart antenna cannot be directly applied, the MIMO technology shall be used to improve the efficiency of power and transmission.

    The efficient controllable modulation and coding technology is the common requirement of future wireless communications. However, due to fixed location, fixed transmission path, and smooth channel fading of wireless transmission between WRs and that between WR and WGW, efficient channel estimation and compensation technology to implement a transmission modulation and coding efficiency are higher than those in a mobile environment to support high-speed communications. Featuring high speed, controllability, manageability and self-adaptation, such technologies as Orthogonal Frequency Division Multiplexing (OFDM), Quadrature Amplitude Modulation (QAM) iteration and Turbo coding/decoding are technologies of first choice.

    The low critical transmit power control technology is key to the improvement of signal power efficiency and is closely related to network topology. With the application of wireless access and dense coverage, a WMN can implement low signal power application. To minimize interference on adjacent WRs, the critical power control of transmit power maximization is very important. Figure 4 is a schematic diagram of critical low-power transmission control. In Figure 4(a), the transmit power is too low, and only partial connection is made; in Figure 4(b), the transmit power is too high, causing excessive mutual crossing and overlapping of the coverage of the individual WRs and serious mutual interference; and in Figure 4(c), the transmit power is controlled to an appropriate critical value, leading to less mutual crossing and overlapping, and all the WRs are connected to the WGW via single hop or multiple hops to implement full connection and best control. Of course, to control the transmit power of a signal, not only network topology shall be considered, such requirements for data service load, transmission delay and Quality of Service (QoS) shall also be considered, to implement the maximum network capacity with excellent comprehensive performance.

1.2 MAC Technology for Multi-Channel Access
Same as the common wireless communications networks, the multi-channel technology providing Media Access Control (MAC) includes Frequency Division Multiple Access (FDMA), Time Division Multiple Access (TDMA), Code Division Multiple Access (CDMA), and Spatial-Division Multiple Access (SDMA) using directional antennas. In actual applications, normally part or combination of all the abovementioned multiple access technologies are applied to form a mutually independent multi-channel access technology without mutual crosstalk. The directional transmission at a fixed point of WR can make full use of the smart antenna technology to implement SDMA and as more independent transmission channels as possible without mutual interference. Therefore, this is another feature of the WMN compared with the traditional wireless communications.

    With support of the multiple access technology, the MAC layer design of a WMN, same as the MAC design of a typical wireless network, is related to access points. The WMN is a multi-hop system instead of a single-hop system, so it needs multi-hop MAC design. First, a nearby Mesh router is selected for access. A WMN is an ad hoc network; and its topology of network route connection and user terminal access varies with geographic location, communications environment, user mobility and WR deployment. As shown in Figure 5, Figure 5(a) is a Mesh topology. The terminal accesses to the Internet access point GW1 to complete the MAC process via three-hop relay. Figure 5(b) shows the same Mesh router and WGW deployment in the same region, without different Mesh topology. The terminal is in the same location. With the same Mesh router, it needs a four-hop relay to access to the Internet access point GW3 to complete the entire MAC process. However, with a different nearby Mesh router, access to GW1 may only need two or three hops. Therefore, the nearby Mesh router selection of a WMN is dynamic, which is different from the normal design. A typical one is the Multi-Channel MAC (MMAC) protocol applied to IEEE 802.11 with the consideration over the interaction between the MAC layer and the network layer and the introduction of Multiple Channel Cooperation Layer (MCCL) to enhance the network capability.

1.3 Routing Technology for Access WGW
The routing technology and related protocols for a user terminal to access to a wireless IP access point via a WR are the most important technologies of a multi-hop WMN. The basic principles for research on and design of the routing technologies and protocols to access to the Internet include the following elements: As less hops as possible, as less delay as possible; as higher data rate as possible; as lower error rate as possible; and as higher routing stability as possible. Thus, special attention shall be paid to the following points during the design of routing protocols for accessing WGW: First, the routing protocol over a WMN shall not make route selection merely according to the “minimum hops”, but instead, comprehensive consideration shall be made over multiple performance measurement indices, and route selection shall be made after comprehensive evaluation; secondly, the routing protocol shall provide network tolerance and robustness support, and can rapidly select a substitute link to avoid service interruption in case of wireless link failure; thirdly, the routing protocol shall be able to utilize the traffic engineering technology to conduct load balance among multiple paths so as to make use of system resources to the greatest extent; and fourthly, the routing protocol shall be able to support routers and user terminals at the same time.

    Typical wireless Mesh routing protocols include Dynamic Source Routing (DSR) Protocol,
Destination-Sequenced Distance-Vector (DSDV) routing protocol, Temporally Ordered Routing Algorithm (TORA) protocol, Ad Hoc on Demand Distance Vector (AODV) protocol, and more. The DSR protocol is the most common peer topology-based reactive Ad hoc routing protocol. It takes active buffer policy, and extracts topologic information from source routes to create routes by comparison. Figure 6 shows a WMN which may involve different uplink/downlink route selection. Normally a Mesh router over a WMN is static, so it in principle does not have power consumption limitation, and does not have any change to router location and routing topology arising from user’s mobility either. Therefore, the existing Ad hoc routing protocol can be simplified, that is, inter-layer design can be made to establish a simpler routing protocol. However, for mobile user terminals, a routing protocol completely similar to Ad hoc shall be used to seek for nearby access points and routes.

    Another problem concerning the routing protocol of an access network is how to make route selection to implement access fairness so that network access opportunity, data rate and communications quality of user terminals are basically consistent. Figure 7 gives the basic route selection mode to implement fairness. In possible conditions, if the basic transfer capabilities of the individual Mesh routers are the same, the parallel access mode should be selected, as shown in Figure 7(a). Thus, the WRs will support their accessed user terminals to connect the Internet respectively at the possible maximum data rate so that different users enjoy the same fairness. When the serial access mode, as shown in Figure 7(b), is taken, if the basic capabilities of the Mesh routers are same, the maximum rate cannot be realized to implement fairness. In such a case, WR4 accesses a user terminal to WGW at a data rate of S4 via WR3. If WR3 already has user access, the rate supported by WR3 is S3-S4. If the maximum capability supported by WR is S, the possible fair access to be implemented is S3=S4=S/2. WR4 does not reach its maximum support capability. WR3 reaches its maximum support capability; however, its direct users can only implement partial access capability of WR3. The serial access mode can relatively implement fair access effectively only when the access capability of WR3 is obviously greater than that of WR4. The consideration over access fairness is also access consideration over the implementation of the maximum capability of the individual Mesh routers over the network so as to maximize the network capacity.

1.4 Wireless Mesh Router Configuration Technology
Network devices normally refer to the Internet access points and Mesh routers. When the coverage is given and if the WGW position can be changed, or when the WR deployment density and number are given, the determination of its location is another basic topic in establishing a WMN. In most cases, the WGW position is determined, so normally research on the configuration of Mesh routers is conducted.

    To implement the maximum network capacity, multi-hop links (routes) in serial configuration are necessary. To improve efficiency, such links operate in time division mode. Therefore, research on time division policy and processing method to avoid collision shall be conducted, as shown in Figure 8. In serial configuration mode, the maximum access capacity of a Mesh router has different requirements in different locations. Different modulation modes and different micro-cell coverage ranges can be used to transfer Mesh routers with heavy tasks. High-speed transmission technology is used to cover a smaller area and reduce the demand of user terminals for direct access. Due to light transfer workload of the Mesh router at multi-hop end position, lower data rate and wider coverage can be used to implement best network capacity, as shown in Figure 9.

2 Typical Applications and Standards
WMN was designed for improving the wireless Internet access and application. Its typical applications cover special regions of a city, streets, interior and exterior of building groups, office areas and homes. The WMN launched by Lam Tech, a British company, to cover special regions of a city, uses 90 Mb/s broadband for the Internet access and four-direction directional antennas for receiving/transmitting. The WMN established by Motorola in Orlando, US is applicable to mobile broadband access, which adopts adaptive transmission, pre-priority MAC and routing protocol; in the IEEE 802.11b WMN established by BelAir (a network company in Canada) on the side of Lake Ontario, each router has directional antennas with three RFs and eight directions, which can dynamically control the signal transmitting power and data rate to implement interior and exterior coverage of buildings with good load balance; and the WMN established by Telabria in the Kent state of Great Britain uses dual carriers compatible with IEEE 802.11 to implement indoor and outdoor coverage of homes or offices. In addition, many other companies also have launched their WMNs and provided broadband data services in different environments, such as Aerial, Firetide, Intel, Microsoft, Nokia, Notel, SkyPilot, and Strix.

    Products and applications launched by these companies are fundamental representatives of the IEEE 802.11, 802.15 and 802.16 standards, concerning related standard recommendations based on wireless Mesh technologies and networks.

    IEEE 802.11s is a wireless distribution system standard, which was proposed in 2004 to expand IEEE 802.11 coverage, led by Intel and Cisco, and was compatible with IEEE 802.11a/b/g. It can be used to implement a multi-hop network which can automatically establish a path, and make self-configuration of topology. Moreover, it can support broadcast and multicast services. IEEE 802.15.1 and 802.15.4 are standards of Bluetooth and Zigbee, both of which involve related recommendations on building WMN. 802.15.1 tends to support wireless low-rate communications in personal surrounding regions and uses simple and convenient hardware to support narrowband multi-hop dispersed networks, while IEEE 802.15.4 adopts Mesh topology to support low-rate communications, so it is more applicable to wireless sensor network applications. IEEE 802.15.5 is a Mesh network topology standard more applicable to Wireless Personal Area Network (WPAN), featuring easy network building. It is used to expand network coverage, reduce router redundancy, and efficiently improve network capability without the increase of transmit power and any influence upon receiving sensitivity.

    IEEE 802.16 is a WiMAX technical standard. To expand special user links, centralized scheduling and distributed scheduling are recommended. IEEE 802.20 and 802.22 are mobile broadband high-speed wireless access system standards involving large-area coverage. They are used in some regions and indoor and outdoor dense application environments, and are technical standards for utilizing WMN.

3 Technology Development and New Applications
Although WMN is a technology which develops at an earlier stage of the Internet application and presently targets at IEEE 802 series wireless access and network applications, its network architecture and networking mode still have high research value and bright application prospects in future wireless mobile communications.

3.1 Mesh Network Architecture in 3GPP LTE
Since the promotion of the standard research on and application of 3G WCDMA, the 3rd Generation Partnership Project (3GPP) organization has made continuous improvement and enhancement of the third generation mobile communications systems based on Wireless Code Division Multiplex Access (WCDMA) and TD-SCDMA. Based on the existing Universal Mobile Telecommunications System (UMTS) R99, 3GPP has successively released the R5 standard version on High Speed Downlink Packet Access (HSDPA) and the R6 standard version on High Speed Uplink Packet Access (HSUPA). Since the 21st century, a representatives of IEEE 802 series, broadband wireless access technologies and standard recommendations have attracted notable attention, particularly, its higher data rate and mobility support gradually pose as competitive edges of the existing mobile communications systems. Therefore, to deal with the technical and market competition, 3GPP initiated the LTE project in 2004. The LTE system uses a bandwidth of 20 MHz, and the peak rate of air interface is 100 Mb/s on the downlink (spectrum efficiency: 5 bps/Hz), and 50 Mb/s on the downlink (spectrum efficiency: 2.5 bps/Hz). With an IP network as the bearer network, to simplify the signaling flow and shorten delay, the LTE has omitted the Node B structure of "RNC + Node B" in UMTS Terrestrial Radio Access Network (UTRAN), which is completely replaced by Enhanced Node B (eNB).

    The topology of the LTE system is shown in Figure 10. IP transmission is applied in the bottom layer between Nodes B. Logically, the X2 interface is used for mutual connection. This is a traditional Mesh network. Such network architecture design is used to support the mobility of user terminals over the entire network and ensure seamless switching of subscribers. However, each Node B is connected to the Service Gateway (SGW) via the S1 interface. The S1 interface also uses the connection mode of Mesh type or partial Mesh type so that a Node B can be connected to multiple SGWs, to provide better support for meeting the ever-increasing demand of telecom carriers and subscribers.

3.2 Mesh Repeater in Wireless Mobile Systems
In mobile communications, it is a normal case to face the access demands in coverage holes (places where signals are too weak to be received stably) between cell edges and some remote areas. In such a case, a repeater is used to serve as a simple Node B extension to meet the coverage needs. Due to their excessive simplicity, poor signal quality, low access capability and strong interference on surroundings, repeaters do not have any upgrade adaptability. In recent years, Mesh repeaters with the introduction of Mesh technologies have been used to replace the traditional repeaters, which is an effective means to achieve the above purpose. As shown in Figure11, Node B provides a limited coverage, so there are weak signal areas between adjacent cells and after tall and large obstacles, but  after the introduction of the Mesh repeater for relay, a perfect coverage emerges. Furthermore, if mobile subscribers change, it can adaptively schedule the operating status of the Mesh repeater and implement reasonable and efficient network capability by operating at the maximum load or suspending its operation based on the actual needs. In this way, the network with Mesh repeaters  has good scalability, reorganization capability, and continuous upgradeability, so that it can provide quality and high-rate broadband services for any subscribers within the coverage.

4 Conclusions
WMN is a kind of network which is developed to meet the wireless Internet access need. It is developed and applied in the traditional wireless broadband access systems. With the rapid development of wireless mobile communications and wide application of the Internet, WMN, as a network technology and network form, has been attached further importance to and promises further development. It will gradually become the fundamental network technology for wireless mobile access networks and penetrate into various future wireless networks.

    WMN grows with the development of Ad hoc networks. Both of them were developed to solve wireless access network problems at the very beginning. However, due to the application of different technical means (Ad hoc networks depend upon mobile subscriber terminals to support routing function and relay access; while WMNs depend upon fixed wireless routers to expand access coverage), their features and applications are obviously different. WMN can be used in applications occasions, which involve personal terminal applications, diversified services, good coverage and public service features. Ad hoc networks are more used in special application occasions featuring complicated environments, strong randomness and great access support change, for example, wireless sensor networks and military networks.

References
[1] SICHITIU M L. Wireless Mesh networks challenges and opportunities [R]. Raleigh, NC, USA: NC State University, USA, 2006.
[2] AKYILDIZ I F, WANG Xudong, WANG Weilin. Wireless Mesh networks: a survey [J]. Computer Networks, 2005, 47 (4): 445-487.
[3] OYMAN O, LANEMAN J N, SANDHU S. Multihop relaying for broadband wireless Mesh networks: from theory to practice [J]. IEEE Communications Magazine, 2007, 45 (11): 116-122.
[4] NGUYEN U T, JIN X. Multicast routing in wireless Mesh networks: Minimum cost trees or shortest path trees? [J]. IEEE Communications Magazine, 2007, 45 (11): 72-77.
[5] SKALLI H, GHOSH S, DAS S K, at　el. Channel assignment strategies for multiradio wireless Mesh networks: Issues and solutions [J]. IEEE Communications Magazine, 2007, 45 (11): 86-95.
[6] WEI H Y, HAAS Z J. Interference-aware IEEE 802.16 WiMAX Mesh networks [C]//Proceedings of 61st Vehicular Technology Conference (VTC2005-Spring): Vol 5 (3102-3106), May 29-Jun 1, 2005, Stockholm, Sweden.

[Abstract] Wireless Mesh Network (WMN) has been actively researched and developed as a new network technology to support broadband high-speed multimedia services. This paper discusses WMN technology and applications, and introduces the basic technologies, typical applications and current development of the WMN. With the future development of wireless distribution technology and network, wireless Mesh technology and network will become the key networking technology and architecture of wireless mobile communications, and will be widely used in various wireless networks with more important roles.