6.4. Cellular NetworksCellular networks use a networked array of transceiver base stations , each located in a cell to cover the networking services in a certain area. Each cell is assigned a small frequency band and is served by a base station. Neighboring cells are assigned different frequencies to avoid interference. However, the transmitted power is low, and frequencies can be reused over cells separated by large distances. The hexagonal pattern of a cell is chosen so that the distance d between the centers of any two adjacent cells becomes the same. Distance d is given by Equation 6.1
where r is the cell radius. A typical practical value for the cell radius is 3 km to 5 km. First-generation cellular networks were analog and used FDMA for channel access. Second-generation cellular networks were digital, using CDMA channel-access techniques. With the advent of the third- and later-generation cellular networks, multimedia and voice applications are supported over wireless networks. A covered area of a cellular network is visualized and approximated by a hexagonal cell served by its own antenna at the center of the hexagon, as shown in Figure 6.7. Figure 6.7. Cellular partitions: (a) real areas; (b) modeled areas
6.4.1. ConnectivityFigure 6.8 shows the connectivity of two mobile users in cellular systems. A cell base station at the center of a cell consists of an antenna , a controller , and a transceiver . The base station is connected to the mobile switching center (MSC). An MSC serves several base stations and is responsible for connecting calls between mobile units. In Figure 6.8, an MSC is connected to two base stations, A and B. An MSC is also connected to the public telephone system to enable communication between a fixed subscriber and mobile subscriber. An MSC also manages mobility and accounting for user billing. Figure 6.8. Connectivity of mobile users in cellular systems.The steps involved in establishing a call between two mobile users in a cellular network are as follows :
The preceding operations use two types of channels for communication between the mobile and a base station, depending on the application of an operation. These two types of channels are control channel and traffic channel . Control channels are used for call setup and maintenance. This channel carries control and signaling information. Traffic channels carry the data between the users. Calls between fixed and mobile subscribers are also possible. An MSC can connect to the public telephone system and enable the connection between a mobile user and a fixed subscriber. 6.4.2. Frequency ReuseThe basic idea of frequency reuse is that if a channel of a certain frequency covers an area, the same frequency can be reused to cover another area. The transmission power of the antenna in a cell is limited to avoid energy from escaping into neighboring cells. We define a reuse cluster of cells as N cells in which no frequencies are identical. Two cochannel cells are then referred to two cells in which a frequency in one cell is reused in the other one. Figure 6.10 shows a frequency-reuse pattern in a cellular network. In this example, each cluster has seven cells, and those cells with the same numbers are cochannel cells. Figure 6.10. Cell clusters and frequency reuse among seven-cell clusters
Let F be the total number of frequencies allotted to a cluster with N cells. Assuming that all cluster cells share an equal number of frequencies, the number of channels (frequencies) each cell can accommodate is Equation 6.2
A cluster can be replicated many times. If k is the number of times a cell is replicated, the total number of channels in the cellular systemso-called system capacity is derived by Equation 6.3
In order to determine the location of a cochannel cell, we start from the center of a cell and move m cells or md km to any direction, then turn 60 degrees counterclockwise, and, finally, move n cells or nd km until we reach a cell with the same frequency. As shown in Figure 6.11, we have a case with m = 3 and n = 1. With a simple geometric manipulation of this situation, we can calculate the distance between the center of the nearest neighboring cochannel cell, D , shown in Figure 6.11: Equation 6.4
Figure 6.11. Nearest cochannel cells
As the distance between the centers of any two adjacent cells is according to Equation (6.1), we can rewrite Equation (6.4) as Equation 6.5
If we connect all the centers of all hexagons for "1" cochannel cellsthe same arguments can be made for 2, 3, ... cochannel cellswe make a larger hexagon, covering N cells with radius D . Knowing that the area of a hexagon is approximately 2.598 x (square of its radius), we can derive the ratio of the areas of the r -radius hexagon, A r , and D -radius hexagon A D as Equation 6.6
Combining Equations (6.5) and (6.6), we calculate the area ratio as Equation 6.7
From the geometry, we can easily verify that the D -radius hexagon can enclose N cells plus 1 / 3 of the cells from each six overlapping peripheral D -radius hexagons. Consequently, the total number of cells covered in a D -radius hexagon is N + 6(1 / 3) N =3 N . Then, , and Equation (6.7) can be simplified to Equation 6.8
This important result gives an expression for the size of the cluster in terms of m and n . As the number of users in a cellular network increases , frequencies allotted to each cell may not be sufficient to properly serve all its users. Several techniques can be used to overcome these problems, as follows:
6.4.3. Local and Regional HandoffsWhen it moves to a new cell, a wireless terminal requests a handoff for a new channel in the new cell. The increase in traffic volume and demand, as well as seamless, high-performance handoff in wireless systems are expected. A successful handoff operation requires certain criteria to be achieved. When a wireless terminal moves from one base-station cell to another, handoff protocols reroute the existing active connections in the new cell. The challenges in wireless networks are to minimize the packet loss and to provide efficient use of network resources while maintaining quality-of-service (QoS) guarantees . Robustness and stability must also be taken into consideration for handoff protocol design. Robustness and stability are especially important when interference or fading in the radio channel results in a request-handoff operation by a wireless terminal. Cellular networks typically have three types of handoffs, as shown in Figure 6.13: channel , cell , and regional . Channel handoff involves transferring a call between channels in a cell. The wireless terminal first initializes a request for a channel change to the base station in a cell, if necessary. If no channels are idle in the cell, the request is rejected, and the wireless terminal has to keep the old channel. Figure 6.13. Cellular handoff types: (a) channel handoff, (b) cell handoff, and (c) regional handoffCell handoff occurs between two adjacent cells. When a wireless terminal moves from one cell to another, the handoff request to the new cell is initialized . If no channels are available in the new cell, the handoff call has to be rejected or terminated . Regional handoff occurs when the mobile user moves from one region to another. From a theoretical standpoint, we can model a handoff process between any two regions, using stochastic models. Consider several hexagonal-shaped regions that consist of a group of hexagonal-shaped cells, as illustrated in Figure 6.14. Since the standard of handoff processes is still being developed, we have to assume that only the boundary cells in a region as labeled in Figure 6.14 are involved the regional handoff model. Similarly, when all channels in the new region are in use, all handoff requests have to be rejected. Figure 6.14. Cellular networks and a regional handoff6.4.4. Mobility ManagementMobility management consists of three functions: location management with user tracking , user authentication , and call routing . Location management involves keeping track of the physical location of users and directing calls to correct locations. Before a user's call is routed to the desired location, the user must be authenticated. Routing involves setting up a path for the data directed to the user and updating this path as the user location changes. In cellular networks, mobile switching centers , in conjunction with base stations, coordinate the routing and location-management functions. The requirement for handoff is dependent on the speed of the mobile terminals and the distance between a mobile user and its cell boundaries. The alternation of states for a mobile unitwhether it is still or mobile while carrying a call in progressalso has to be considered for handoff analysis. In reality, a handoff is needed in two situations: (1) the signal strength within the cell site is low or (2) when a vehicle reaches the cell boundary. We assume that the handoff model is free of signal-strength obstructions. Our second assumption prior to reaching a cell boundary is that a vehicle with a call in-progress alternates between still (stop state) and moving (go-state). Stop-and-Go ModelThe alternation of stop-and-go states can be modeled, using a simple state machine. In state 0 (stop), a vehicle is at rest but has a call in progress. In state 1 (go), the vehicle moves with an average speed of k mph and has a call in progress. Let ± i , j be the rate at which a system with i states moves from state i to state j . For example, ± 0,1 is the rate at which state 0 leaves for state 1, and ± 1,0 is the rate at which state 1 leaves for state 0. In our case, it is clear that ± 0,0 =- ± 0,1 and that ± 1,1 =- ± 1,0 . The time in the stop state is an exponential random variable with mean 1 /± 0,1 (see Appendix C). The time in the go state also is an exponential random variable, with mean 1 /± 1,0 . Let P i (t) be the probability that a vehicle having a call in progress is in state i at time t . According to the Chapman-Kolmogorov theory on continuous-time Markov chain (explained in Section C.5.1), we can derive Equation 6.9
where is the time differential of relative state j probability. Applying Equation (6.9) for a system with two states, 0 and 1, we have Equation 6.10
Equation 6.11
where P ( t ) and P 1 ( t ) are the probability that a vehicle having a call in progress is in states 0 and 1, respectively, at time t . Knowing that the sum of probabilities is always 1, we get P ( t ) + P 1 ( t ) = 1. This equation and Equation (6.10) can be combined to form a first-order differential equation: Equation 6.12
The total solution to the differential equation consists of a homogeneous solution, P h (t) , plus a particular solution, P p (t) . Knowing that , where P h (0)= P (0), we can obtain the general solution for Equation (6.12) by Equation 6.13
where P ( t ) is the probability that a vehicle with a call in progress is in state 0 at time t . Similarly, we obtain the general solution for Equation (6.11) by Equation 6.14
where P 1 ( t ) is the probability that a vehicle with a call in progress is in state 1 at time t . There are four cases for a vehicle to change states.
Now, consider the mobilized model shown in Figure 6.15. To show the probability of a call to reach a cell boundary with an average speed k m/h with stop and go or the probability of requiring a handoff, let s be the average speed of the vehicle, where s = 0 in state 0 and s = k in state 1. Let x(t ) be the vehicle's position at time t , assuming x (0) = 0, and let d b be the distance a vehicle takes to reach a cell boundary. Suppose that t is a random variable representing a channel holding time, or the time a vehicle takes to reach a cell boundary. Let P i (t , d b ) be the probability that a vehicle in state i with a call in progress is at the cell boundary at time t , where i ˆˆ{0, 1}. Consequently, the probability that a vehicle reaches a cell boundary with speed s undergoing stop-and-go states is Figure 6.15. Cellular-handoff model with mobility
Equation 6.15
where Equation 6.16
and Equation 6.17
In case 1, since a vehicle is resting all the time, with an average speed of 0 m/h, the probability of reaching a cell boundary is clearly 0 percent. In contrast, for a vehicle moving with an average speed ( k ) (case 2), the chance of reaching a cell boundary is always 100 percent. Thus, when a vehicle is either at rest or moving, the probability of requiring a handoff is independent of d b . 6.4.5. Generations of Cellular SystemsFirst-generation cellular systems were mostly analog. Channels were allotted to a single user, and each user had dedicated access to the channel. This led to underutilization of resources. Second-generation systems were digital and supported higher data rates, providing digital traffic channels and digitized voice before transmission over the channel. Data digitization made it simple to implement an encryption scheme. The digital traffic also made it possible to deploy better error detection and correction. Finally, multiple users shared a channel by using multiple-access schemes, such as TDMA or CDMA. Third and later generations of wireless networks provide high data rates and support multimedia communications, in addition to voice communications. The main objectives for these cellular networks are to achieve quality voice communications, higher data rates for stationary and mobile users, support for a wide variety of mobile devices, and adapting to new services and technology usable in a wide variety of environments, such as offices, cities, and airplanes. Design issues involved the design of CDMA-based systems are channel-usage bandwidth limitation (5 MHz), chip rate , and multirate capability to provide different data rates on different channels for each user. The multirate scheme can scale effectively to support multiple applications from each user. 6.4.6. CDMA-Based Mobile WirelessThe most commonly used second-generation CDMA scheme is IS-95, which consists of two components : a forward link and a reverse link . The forward link consists of 64 CDMA channels, each operating at a bandwidth of 1.288 MHz. The channels are of four types:
If the data rate is low, bits are replicated to increase the rate to 19.2 Kb/s. This process, called symbol repetition , is followed by a scrambling process for increasing privacy and reducing the interference between users. The next step is a process to control the power output of the transmitting antenna. The data is then processed , using direct-sequence spread spectrum (DSSS) (see Chapter 4). This process spreads the transmitted signal from 19.2 Kb/s to 1.288 Mb/s. The Walsh matrix is used to generate the pseudorandom sequence. The signal is then modulated with the QPSK modulation scheme being transmitted onto the medium. The IS-95 reverse link consists of up to 94 CDMA channels, each operating at 1.228 MHz. The link supports up to 32 access channels and 62 traffic channels. The access channels are used for call setup, location update, and paging. The reverse-link transmission process is very similar to forward-link transmission. The convolution encoder has a rate of 0.333. Thus, the data rate is tripled, to 3 x 9.6 = 28.8 Kb/s, followed by the data-scrambling process achieved by block interleaving. In the reverse link, the Walsh matrix is used to increase the data rate to 307.2 Kb/s as the data from the block interleaver is divided into 6-bit units to improve reception at the receiver. The data-burst randomizer is used in conjunction with the long code mask to reduce interference from other users. In the case of the reverse link, the long code mask is unique for each mobile unit. The resultant data is modulated, using the orthogonal QPSK modulator (OQPSK). The OQPSK scheme is used because the spreading codes need not be orthogonal for the reverse link.
|