Wimax standards and Security The Wimax

Petar Djukic and Shahrokh Valaee^∗


CONTENTS

1. 
   Introduction 147
   
2. 
   802.16 Time Division Multiple Access 150
   
   1. 
      802.16 Physical Layer 150
      
   2. 
      TDMA Framing and Transmission Timing 151
      
   3. 
      Transmission Scheduling in the Logical Channels 154
      
      1. 
         The Basic Channel 154
         
      2. 
         Distributed Election Scheduling Broadcast
         

Channels 154

3. 
   Tree-Based Scheduling Broadcast Channels 157
   
4. 
   Best Effort Broadcast Channel 158
   
5. 
   Transmission Scheduling in the Data Channels 158
   
1. 
   Network Entry and Synchronization 161
   
1. 
   802.16 Mesh Networking 162
   
   1. 
      802.16 MAC Connections 162
      
   2. 
      Mesh Network Addressing 164
      
   3. 
      QoS-Aware Convergence Sublayer 165
      
2. 
   Network Security 168
   
   1. 
      Network Authentication 168
      
   2. 
      Backbone Hop-by-Hop Security 169
      
   3. 
      User End-to-End Security 170
      
3. 
   Conclusion 171
   

References 171
8.1 Introduction
Wireless mesh networks interconnect access points (APs) spread out over a large geographical area. Wireless terminals (WTs) connect to the APs on
^∗ This work was sponsored in part by LG Electronics Corporation.
147



Mobile user Static users


FIGURE 8.1
A mesh network has a number of static backbone nodes that carry traffic for users in the network. Each WT connects to an AP at the edge of the network, and this AP sends WT’s traffic over the backbone to the point-of-presence, which is connected to the Internet. Since there is only one high-speed Internet connection for many APs, the network has a low operational cost.

their first hop. Then, their traffic is carried by the wireless mesh to the point- of-presence (POP) where it can go to the Internet (Figure 8.1). The POP is the only node in the network connected to the Internet and can act as a base station (mesh coordinator). In urban areas, mesh networks interconnect wireless hot spots. Mesh networks decrease the cost of running the hot spots since they only require a single POP broadband connection for the whole network. For example, using a mesh network to interconnect 133 existing hot spots in the Toronto downtown area would decrease the total cost of running the hot spots by 70% [1]. Mesh networks can also be used to provide the wireless last mile in rural areas where it is impractical to provide wired connectivity due to sparseness of customers. This is the idea behind rooftop networks [2], where each house has a mesh node connecting it to neighboring houses while providing wireless access to the devices in the house.

Current mesh networks use 802.11 technology to interconnect the mesh
backbone [3,4]. However, 802.11 technology is a decade old and was not designed for mesh networks. In particular, 802.11 lacks the extensions to provide quality-of-service (QoS) in multihop wireless environments [5]. The
802.11 protocol also lacks security extensions needed to provide WTs with pri- vacy and security across the mesh backbone. These problems are addressed by the 802.16 mesh technology [6]. IEEE 802.16 uses time division multiple access (TDMA) technology to provide QoS and encryption for security and privacy. This chapter reviews 802.16 mesh technology and proposes solutions needed in the network layer to take advantage of 802.16 mesh extensions.
IEEE 802.16 mesh uses TDMA technology to provide link-level QoS in the network. In TDMA, QoS required by WTs is negotiated in terms of

end-to-end bandwidth reserved for each WT on links connecting it to the POP. QoS is enforced at each link with scheduled access to the wireless chan- nel. Link bandwidth is allocated over frames with a fixed number of slots and a scheduler assigns slots to links. During each slot, a number of links that do not conflict with each other may transmit simultaneously. Two links conflict with each other if transmissions by one link prevent packet reception at the other. The bandwidth of each link is given by the number of slots assigned to it in the frame and the modulation used in the slots.

The 802.16 mesh protocol specifies two scheduling protocols for assign- ment of link bandwidths: centralized and decentralized scheduling protocols. The centralized scheduling protocol is used by the base station (mesh coor- dinator) to establish network-wide schedules. In contrast, the decentralized scheduling protocol is used to negotiate pairwise bandwidth assignments between mesh routers. The centralized scheduling protocol can be used to establish network-wide end-to-end QoS; however, the decentralized scheduling protocol is not expected to establish end-to-end QoS.
In 802.16, links between routers are managed with logical connections. Logical connections are established between mesh routers within the wire- less range of each other and remain valid as long as the network operates. However, a connection may be inactive if it is not assigned any TDMA slots. Using a connection-oriented protocol is appropriate for mesh net- works since mesh routers are usually static with respect to each other. The connection-oriented nature of 802.16 protocol significantly improves the efficiency of the mesh. For example, the protocol uses a combination of an 8-bit network ID 16-bit mesh ID, and an 8-bit link ID to associate transmissions with links, compared to 48-bit Ethernet address pairs used by 802.11.
Since 802.16 is a connection-oriented protocol, the network stack used on
802.11 mesh nodes is not applicable for 802.16 networks for several reasons. First, 802.16 mesh networks do not have layer-2 broadcast capabilities and use a convergence sublayer (CS) to multiplex Internet protocol (IP) packets to connections. Therefore ARP [7] is not needed. Second, when a medium access control (MAC) layer scheduling algorithm changes the state of a connection, the routing protocol used on the node should be notified of the change so that routes can be adjusted accordingly. A change in link status may propagate routing changes, which affects QoS. It is therefore necessary to design a net- work layer that is aware of the TDMA nature of 802.16 networks. Third, since
802.16 mesh networks are intended for infrastructure-based mesh networks, the 802.16 routers are static and always on, meaning that the connection life- time is in the order of the network lifetime. The scale of the connection lifetime makes it possible to establish hop-by-hop security in the mesh backbone, by keeping a private key in sync on both sides of the connection. In 802.16, private keys are distributed and managed with a key management protocol initiated by the base station.
This chapter reviews the networking aspects of 802.16 mesh networks with a focus on exposing scheduling, routing, and security problems in

the protocol. We describe the current state-of-the-art research addressing the problems, and we propose our solutions to some of the problems left open in the standard. Section 8.2 presents an overview of TDMA technology used in 802.16 mesh networks and the scheduling algorithms required to man- age TDMA slots. We review the current research analyzing the scheduling algorithms provided in the standard. We also review the research proposing scheduling algorithms required by the standard but left open to the imple- mentation. Section 8.3 presents an overview of the network layer architecture in 802.16 mesh networks, including routing and addressing issues introduced by TDMA technology. The 802.16 standard specifies that the IP layer should be connected to the 802.16 hardware with a CS; however, the implementa- tion details of the CS are left out. We specify a CS that takes advantage of QoS inherently available in 802.16 mesh protocol and integrates it with IP DiffServe architecture [8]. Section 8.4 presents an overview of the security architecture in 802.16 mesh networks and the research exposing the secu- rity flaws in the standard. We present our security additions, which enhance end-to-end security in the network layer.

2.