Cellular Networks
WiFi can provide high speed connectivity at low cost, but its coverage is limited to within a home or office building. In contrast, cellular networks are designed to provide wide area coverage to both static and mobile users. Cellular network is the oldest communications network technology, which has now gone through several generations of evolution. In this chapter, we shall first learn the fundamentals concepts of cellular networks before examining the advancements brough forth by each generation.
7.1 Beginning of Cellular Networks
Back in 1968, AT&T Bell Labs submitted a plan [Rappaport2002] to FCC that they could provide radio communication services to the entire nation with limited spectrum by dividing the spectrum into several frequency bands and then allocating them in hexagonal cells as illustrated in Figure 7.1. Using this pattern, no two adjacent cells would be using the same frequency band, making it possible to cover the entire nation with the limited frequency bands and still avoid interference.
Copyright By PowCoder代写 加微信 powcoder
Figure 7.1 Reusing 7 frequencies to cover a large area with hexagonal cells. No two adjacent cells use the same frequency.
7.2 Initial Deployments of Cellular Systems in the US
In 1981, FCC set aside a total of 40MHz in 800MHz spectrum for cellular licensing [FCC800MHz]. For the initial deployment of cellular systems in US, the whole country was divided into 734 areas called Cellular Market Areas (CMAs). To avoid monopoly, it was decided that every CMA would be covered by two competing carriers, A and B. B stands for Bell, and A represents the alternate.
© 2022 [Wireless and Mobile Networking, CRC Press 2022]
In the initial cellular deployments, uplink and downlink frequencies were different to avoid interference between transmit and receive antennas on the same device, i.e., Frequency Division Duplexing (FDD) was used. With FDD, a pair of frequencies therefore were needed to support a call. Each uplink/downlink channel was allocated 30kHz for a total of 60kHz for the duplex voice call [Rappaport2002].
7.3 Cell Sites
Cellular systems need to install radio towers (base stations) to transmit and receive calls. Where should they put the tower. In the beginning they were building towers from scratch in some places, which is very costly. Then the carriers wanted to use existing infrastructure, but due to wireless radiation as well as pollution of scenery, no one wanted cell towers near their house. There is this acronym NIMBY (not in my backyard). Finally, carrier started to install towers on roof tops of schools, churches, hotels, etc. as well on traffic lights, streetlamps and so on for a fee to the owners. For non-profit organizations, such as schools and churches, there was a great way of making money. Even some fake trees were planted to hide base stations as shown in Figure 7.2.
Figure 7.2 Base stations are erected on top of many different objects.
Fixed tower sites are good for most of the time, but they cannot handle sudden increase in demand in a given area. To serve a sudden surge of people in a given area, such as a big circus or fair, the operators bring CoWs or Cell on Wheels. The whole base station is fitted on top of a van, so the van can go anywhere where there is a demand as shown in Figure 7.3.
© 2022 [Wireless and Mobile Networking, CRC Press 2022]
Figure 7.3 Cell on Wheels
7.4 Macro, Micro, Pico, Femto Cells
As the population started to grow, a single cell tower could not connect all users. So they started to deploy different sizes of cells to meet the demand. There are four different sizes of cell, Macro, Micro, Pico, and Femto, as shown in Figure 7.4.
Macro are the normal original cells with roughly 1Km radius. Micro covers a neighborhood of less than 1Km. Pico cells are deployed in busy public areas, such as malls, airports, etc., covering an area of about 200 m. Finally, femto cells are installed inside a home or office covering 10m to provide good coverage (strong signals). Some operators provide femto cells for free to attract and retain customers.
Figure 7.4 Macro, Micro, Pico, and Femto cells.
7.5 Cell geometry
Although there is no regular cell geometry in practice due to natural obstacles to radio propagations, a model is required for planning and evaluation purposes. A simple model would be for all cells to have identical geometry and tessellate perfectly to
© 2022 [Wireless and Mobile Networking, CRC Press 2022]
avoid any coverage gaps in the service area. Radio propagation models lead to circular cells, but unfortunately circles do not tessellate!
As shown in Figure 7.5, three options for tessellation are considered: equilateral triangle, square, and regular hexagon. Hexagon has the largest area among the three; hence it is typically used for modelling cellular networks.
Figure 7.5 Tessellating cell shapes.
7.6 Frequency reuse and clustering
Earlier we discussed how AT&T proposed to cover the entire nation by simply reusing only 7 frequencies. Frequency reuse is possible because the signal from the cell tower gets weak at the cell border and hence loses its capacity to interfere with other communications happening far away from the current cell.
To keep the interference to a minimum, it’s a common practice to avoid using the same frequency in adjacent cells. To achieve this, all cells in the service area are grouped into many clusters. The total spectrum is then divided into sub-bands that are distributed among the cells within a cluster in such a way that two adjacent cells do not share the same sub-band. Figure 7.6 shows examples of clusters sizes of 4, 7 and 19 where N represents the cluster size and the cluster borders are shown with solid black lines.
© 2022 [Wireless and Mobile Networking, CRC Press 2022]
Figure 7.6 Frequency reuse with different cluster sizes.
7.7 Characterizing Frequency Reuse
The next question we want to answer is: how much reuse, or what is the extent of frequency reuse, can we achieve for a given cluster size? Let us do the mathematics and find out. Let us assume the following notations, which are also illustrated in Figure 7.7:
D = minimum distance between centers of cells that use the same band of frequencies (a.k.a. co-channel cells)
R = radius of a cell
d = distance between centers of adjacent cells. Note that d<2R due to the overlapping of cells, which enables seamless handover from cell to cell for a mobile user. The exact value is d = RÖ3.
N = number of cells in repetitious pattern (Cluster), also called the reuse factor1; note that each cell in the cluster uses unique band of frequencies
For hexagonal cell pattern, N cannot assume arbitrary numbers. It is rather given by the following formula [Rappaport2002]:
N = I2 + J2 + (I x J), where I, J = 0, 1, 2, 3, ...
Therefore, the possible values of N are 1, 3, 4, 7, 9, 12, 13, 16, 19, 21, and so on. Note that some values are not possible. For example, we cannot have a cluster size of 5, because there are no combinations of integers, or I and J, that will provide 5.
1 Sometimes, the reuse factor is represented by the fraction 1/N.
© 2022 [Wireless and Mobile Networking, CRC Press 2022]
Finally, D/R is called the reuse ratio. From hexagonal geometry, it can be shown that that 𝐷⁄𝑅 = √3𝑁, which means 𝐷⁄𝑑 = √𝑁.
Example 7.1
What would be the minimum distance between the centers of two cells with the same band of frequencies if cell radius is 1 km and the reuse factor is 1/12?
R = 1 km, N = 12
D = (3x12)1/2 1 km = 6 km
Figure 7.7 Illustration of frequency reuse notations.
7.8 Locating co-channel cells
Given a tessellated hexagonal cellular pattern of cluster size N, can we identify the co-channel cells? Yes, we can do this using the following simple rule.
First, obtain the I and J values that make up N. For example, for N=4, we could have I=0 and J=2, or I=2 and J=0. Now, to identify the cochannel cell of a particular cell A, move I cells in any direction from the centre of A, turn 60◦ counterclockwise and then move J cells. Note that there are 6 possible directions in a hexagonal cell, each separated by 60 degrees from their neighbours as illustrated in Figure 7.8. For N=19, Figure 7.9 illustrates the rule for finding the cochannel cells in a hexagonal cellular network.
© 2022 [Wireless and Mobile Networking, CRC Press 2022]
Figure 7.8 Six directions of a hexagon
Figure 7.9 Locating co-channel cells in hexagonal cellular network: N=19 (I=3, J=2).
7.9 Spectrum distribution within cell cluster
We have learned that the spectrum available to a cellular operator can only be reused outside the cluster. Next, we are going to examine how the spectrum is distributed among the cells within the cluster.
For simplicity, it is assumed that the total spectrum is divided equally among all cells in the cluster. Let T denote the total number of available channels, N the cluster size, and K the number of channels per cell. Then we have K = T/N.
Cells are usually divided into sectors where a frequency received in one sector may not be received in another. Channels allocated to a cell is then further sub-allocated to different sectors according to the load or demand in each sector. Sectorized allocation of channels can also help minimize inter-cell interference, which is a major issue arising from spatial reuse of spectrum in cellular networks. For example, the cellular network in Figure 7.10 can reuse its spectrum with a cluster size of only 1 as two adjacent cell sectors do not use the same frequency.
© 2022 [Wireless and Mobile Networking, CRC Press 2022]
7.10 Frequency Reuse Notation
To describe a given frequency reuse pattern for sectorized cells, we can use a notation like NxSxK, where N is the number of cells per cluster (cluster size), S is the number of sectors in a cell, and K is the number of frequency allocations per cell.
Figure 7.9 is an example of a 1x3x3 frequency reuse pattern where different colors represent different frequencies. Note that in this example, the same frequencies are used in every cell (N=1). There are three sectors (S=3) in each cell, which is shown by dotted line dividing each cell into three equal geographical regions. Three frequencies have been allocated per cell (K=3).
In Figure 7.9, each sector uses one frequency, but in real life, multiple frequencies may be allocated to a given sector (heavily populated), while other sectors may have just one or even no frequency allocated. Here K=3 only means that three frequencies are allocated per cell, but how the frequencies are distributed between the sectors is not captured by the NxSxK notation.
Figure 7.10 An example of a frequency reuse with cluster size of 1 using sectorized antenna.
Figure 7.11 shows 6 more examples of frequency reuse notations. In this figure, frequencies are shown as numbers within the sector. Again, frequencies are evenly distributed among the sectors. For example, if a cell is allocated 3 frequencies, each sector is allocated a different frequency.
Figure 7.11 also shows the location of a subscriber station (SS) within the cell and which tower it is likely to get the signal from using a red arrow from tower to the SS. There is a red arrow from a tower if it is towards the sector with the same frequency. Let us examine how the SS receives signal from different towers by following the red arrows in the figure for different patterns.
© 2022 [Wireless and Mobile Networking, CRC Press 2022]
For the first pattern (1x3x1), there is only one frequency. Given the current location of the SS shown in the figure, it can receive the same frequency signal from five other cells besides the current cell. Similarly, for 3x3x1, the SS is receiving frequency 2, so it will receive frequency 2 signal from 4 other cells. Fortunately, if the SS is located in the center of the cell, then the signal from the current cell tower will be the strongest, which will help it to connect to this tower without any confusion.
A problem arises when the SS is located close to the edge. If it receives the same frequency signals from both the cells, then the signal strengths may be close to each other creating confusion. This leads to a so-called ping-pong effect, where the SS may switch between towers as it moves.
Figure 7.11 Examples of 6 different frequency reuse patterns
7.11 Fractional Frequency Reuse
The cell-edge problem can be addressed by a concept called fractional frequency reuse, which controls the signals strengths of the frequencies in a way such that some frequencies can only be heard in the center (not heard in the edge) while only one of them can be heard in the edge. This is shown in Figure 7.12. We can see that with fractional frequency reuse concept, stations in the edge no more has the confusion because no border uses the same frequency.
© 2022 [Wireless and Mobile Networking, CRC Press 2022]
7.12 Handoff
Figure 7.12 Fractional frequency reuse
User mobility poses challenges for cellular networks. As the user starts to leave the coverage of a cell, the RSS becomes too weak. The user then must connect to a new BS with a stronger RSS to keep the connection to the network. Disconnecting from one and connecting to a new BS during an on-going session is called handoff, which is illustrated in Figure 7.13.
Figure 7.13 The handoff process in cellular networks.
To handoff successfully, the new BS must have available channels to support the on- going call; otherwise the call will be dropped. Dropping an ongoing call is worse than rejecting a new call. BSs therefore usually reserve some channels, called guard channels, exclusively for supporting handoff calls. Unfortunately, guard channels increase the blocking probability of new calls. The number of guard channels is left to the operators to optimize, i.e., it is not part of the standard.
Example 7.2
© 2022 [Wireless and Mobile Networking, CRC Press 2022]
A particular cellular system has the following characteristics: cluster size=7, uniform cell size, user density=100 users/sq km, allocated frequency spectrum = 900-949 MHz, bit rate required per user = 10 kbps uplink and 10 kbps downlink, and modulation code rate = 1 bps/Hz. How many users per cell can be supported and what cell sizes are required?
49 MHz/7 = 7 MHz/cell; For symmetric bandwidth requirement in uplink/downlink, we have 3.5 MHz/uplink or downlink
10 kbps/user = 10kHz/user (1bps/Hz); users/cell =3.5MHz/10kHz = 350
100 users/km2; to connect 350 users, the cell area has to be 350/100= 3.5 km2
πr2 = 3.5; r = 1.056 km Example 7.3
A particular cellular system has the following characteristics: cluster size=7, uniform cell size, user density=100 users/sq km, allocated frequency spectrum = 900-949 MHz, bit rate required per user = 10 kbps uplink and 10 kbps downlink, and modulation code rate = 1 bps/Hz. If the available spectrum for uplink/downlink is divided in to 35 channels and TDMA is employed within each channel:
1. What is the bandwidth and data rate per channel?
2. How many time slots are needed in a TDMA frame to support the required number of users?
3. If the TDMA frame is 10ms, how long is each user slot in the frame?
4. How many bits are transmitted in each time slot?
1. 49 MHz/7 = 7 MHz/cell; For symmetric bandwidth requirement in uplink/downlink, we have 3.5 MHz/uplink or downlink
3.5 MHz/35 = 100 kHz/Channel = 100 kbps per channel
2. With 10 kbps/user, we have 10 users/channel
3. 10 ms/10 = 1ms
4. 1 ms x 100 kbps = 100 b/slot
7.13 Cellular Telephony Generations
As we have discussed, cellular telephony started back in 1980s. That was called the first generation of cellular networks. Since then, the technology continued to evolve to meet the demand in terms of number of people and devices that want to connect as well as the nature of traffic they want to send, such as voice vs. data.
In cellular word, the major changes are marked as a generation (G), which roughly lasts for 10 years. Any major changes in between the 10 years is then marked as fraction of 10, such as 2.5G. Figure 7.14 shows the evolution of these generations. The figure shows how the evolution in terms of standardization is happening in the
© 2022 [Wireless and Mobile Networking, CRC Press 2022]
US (or North America) and in Europe in the core technology, such as Analog vs. Digital, and in traffic types, such as voice vs. data.
Following are some of the key points to note:
Technology and Traffic: The first generation (1G) was analog and using FDMA to transmit only voice. It started digital transmission starting from 2G, but it was voice. Data could only be transmitted by converting it to voice signals using modems. Actual data transmission started from 2.5G and by now it is mostly data. Voice is now transmitted over data services.
Standardization in North America and Europe: North America and Europe continued using different standards until the end of 3G, when they converged to LTE (Long Term Evolution).
Figure 7.14 Evolution of cellular telephony
Table 13.1 shows more details of each generation. Most of the standards, such as GPRS, EDGE, WCDMA and so on are now almost extinct. One standard that survived and very much in use worldwide including in North America is GSM.
Table 7.1 Cellular generations Generation Traffic Standards
2G - 1990 Digital Voice. TDMA • cdmaOne: Qualcomm. International © 2022 [Wireless and Mobile Networking, CRC Press 2022]
1G – 1980s
Analog Voice. FDMA
• AMPS: Advanced Mobile Phone System
• TACS: Total Access Communications
Standard IS-95.
• Digital AMPS (D-AMPS)
• GSM: Global System for Mobile
Communications
2.5G - 1995
Voice+data
• 1xEV-DO: Evolution Data Optimized
• 1xEV-DV: Evolution Data and Voice
• General Packet Radio Service (GPRS)
• Enhanced Data Rate for GSM Evolution
Voice+High-speed data. All CDMA
• CDMA2000: Qualcomm. International Standard IS-2000.
• W-CDMA: Wideband CDMA
• TD-SCDMA: Time Division
Synchronous Code Division Multiple
Access (Chinese 3G)
• 384 kbps to 2 Mbps
Voice+Higher-speed Data
• EDGE Evolution
• High-Speed Packet Access (HSPA)
• Evolved HSPA (HSPA+)
• Ultra Mobile Broadband (UMB)
High-speed data+VOIP. OFDMA
• WiMAX 16e (Worldwide Interoperability for Microwave Access)
• Long Term Evolution (LTE)
Very High-speed Data
• WiMAX 16m or WiMAX2 • LTE-Advanced
• 100 Mbps – 1 Gbps
Ultra High-speed data + Ultra Low Latency + Massive connectivity
GSM stands for Global System for Mobile Communications. It is now implemented in most cell phones world-wide and most countries are using GSM. A phone without GSM support therefore would not do much.
The interesting thing is that GSM was designed back in1990. Three decades on, it is still a very popular technology. GSM uses Time-Division Multiple Access (TDMA) instead of Frequency Division Multiple Access (FDMA) used in 1G. Figure 7.15
© 2022 [Wireless and Mobile Networking, CRC Press 2022]
shows the difference between FDMA and TDMA. In FDMA, once a frequency is allocated to a user, no one else was allowed to use that frequency. This wasted a lot of system capacity. With TDMA, the same frequency could be used by multiple users shared in time. This is possible because there are many silence periods in voice communication, which can used for other users.
GSM is defined for all major frequency bands used throughout the world. Specifically, it supports the four bands, 850/900/1800/1900 MHz, hence called quad- band. Handsets not supporting quad-band may not operate in some countries.
The biggest invention of GSM was to separate the user from the handset. Prior to GSM, user subscript information was tied to the handset hardware. It made it di
程序代写 CS代考 加微信: powcoder QQ: 1823890830 Email: powcoder@163.com