Making Connections Efficient: Multiplexing and Compression

Transcription

1 Fundamentals of Networking and Data Communications, Sixth Edition 5-1 Making Connections Efficient: Multiplexing and Compression Chapter 5 Learning Objectives After reading this chapter, students should be able to: Describe frequency division multiplexing and list its applications, advantages, and disadvantages Describe synchronous time division multiplexing and list its applications, advantages, and disadvantages Outline the basic multiplexing characteristics of T-1 and SONET/SDH telephone systems Describe statistical time division multiplexing and list its applications, advantages, and disadvantages Cite the main characteristics of wavelength division multiplexing and its advantages and disadvantages Describe the basic characteristics of discrete multitone Cite the main characteristics of code division multiplexing and its advantages and disadvantages Apply a multiplexing technique to a typical business situation Describe the difference between lossy and lossless compression Describe the basic operation of run-length, JPEG, and MP3 compression Chapter Outline 1. Introduction 2. Frequency Division Multiplexing 3. Time Division Multiplexing a. Synchronous time division multiplexing T-1 Multiplexing SONET/SDH Multiplexing b. Statistical time division multiplexing 4. Wavelength Division Multiplexing

2 Fundamentals of Networking and Data Communications, Sixth Edition Discrete Multitone 6. Code Division Multiplexing 7. Comparison of Multiplexing Techniques 8. Compression Lossless versus Lossy a. Lossless compression b. Lossly compression 9. Business Multiplexing In Action 10. Summary Lecture Notes Introduction Under the simplest conditions, a medium can carry only one signal at any moment in time. Many times, however, we want a medium to carry multiple signals at the same time. This technique of transmitting multiple signals over a single medium is multiplexing. Multiplexing is a technique performed at the physical layer of the OSI model or the interface layer of the Internet model. Frequency Division Multiplexing Frequency division multiplexing is the assignment of non-overlapping frequency ranges to each user of a medium. So that multiple users can share a single medium, each user is assigned a channel. A channel is an assigned set of frequencies that is used to transmit the user s signal. Time Division Multiplexing Time division multiplexing directly supports digital signals. In time division multiplexing, sharing of the signal is accomplished by dividing available transmission time on a medium among users. Since time division multiplexing was introduced in 1960s, it has split into two roughly parallel but separate technologies: synchronous time division multiplexing and statistical time division multiplexing.

3 Fundamentals of Networking and Data Communications, Sixth Edition 5-3 Wavelength Division Multiplexing Wavelength division multiplexing is a fairly new technology that multiplexes multiple data streams onto a single fiber optic line. Unlike frequency division multiplexing, which assigns input sources to separate sets of frequencies, and time division multiplexing, which divides input sources by time, wave division multiplexing uses different wavelength lasers to transmit multiple signals. Discrete Multitone Discrete multitone is a multiplexing technique commonly found in digital subscriber line (DSL) systems. Each subchannel in DSL performs its own quadrature amplitude modulation. Up to 256 subchannels can be combined to produce one downstream path. Code Division Multiplexing Code division multiplexing is a relatively new technology most commonly found in cellular telephones which allows multiple mobile devices to transmit at the same frequencies and at the same time. In order to separate the signals, each mobile device is assigned a unique binary code. When the devices transmits a binary 1, the unique binary code is transmitted instead. To transmit a binary 0, the mobile device would send the inverse of the unique binary code. Comparison of Multiplexing Techniques Frequency division multiplexing relies on analog signaling and is the simplest and most noisy of all the multiplexing techniques. Synchronous time division multiplexing is also relatively straight forward, and like frequency division multiplexing, input devices that have nothing to transmit can waste transmission space. The big advantage of synchronous TDM is the lower noise during transmission. Statistical TDM is one step above synchronous TDM because it transmits data only from those input devices that have data to transmit. Thus, statistical TDM wastes less bandwidth on the transmission link. Dense wavelength division multiplexing is a very good, albeit expensive, technique for transmitting multiple concurrent signals over a fiber optic line. Compression Lossless versus Lossy The compressing of data can be either lossless (in which no data is lost) or lossy (in which some data is lost). Common lossless compression techniques include run-length encoding and the Lempel-Ziv techniques. Typical lossy compression techniques include MP3, JPEG, and MPEG.

4 Fundamentals of Networking and Data Communications, Sixth Edition 5-4 Business Multiplexing In Action The In Action example for this chapter involves the XYZ Corporation which has two buildings, A and B, separated by a distance of 300 meters. A 3-inch diameter tunnel runs underground between the two buildings. Building B contains 66 text-based terminals that need to be connected to a mainframe computer in Building A. The text-based terminals transmit data at 9600 bits per second. What are some good ways to connect the terminals in Building B to the mainframe computer in Building A? Quick Quiz 1. What types of applications might use frequency division multiplexing? Time division multiplexing? Dense wavelength division multiplexing? FDM: TV, radio, cellphones; TDM: PCS, terminal-mainframe connections, telephone systems; DWDM: long-haul telecommunications 2. What is the primary advantage of TDM over FDM? TDM can use digital signals which are better for removing noise. 3. Where is discrete multitone commonly used? It is commonly used in cellular telephone systems. 4. How does code division multiplexing work? Each mobile device is assigned a unique binary code for transmitting 1s and 0s. Discussion Topics 1. Is frequency division multiplexing going to disappear in the near future? 2. Will wavelength division multiplexing revolutionize multiplexing? 3. Is there a practical limit to how many devices can transmit simultaneously using code division multiplexing? 4. Do people actually notice a loss of signal when MP3 or JPEG is incorporated?

5 Fundamentals of Networking and Data Communications, Sixth Edition 5-5 Teaching Tips 1. Compare FDM to a multiple lane highway. Many concurrent lanes of traffic, with the stripe down the road acting as the guardbands. 2. Mention the recent development of the blue laser (and the blue LCD) and its effect on dense wavelength division multiplexing (and other fields of technology). Solutions to Review Questions 1. What are the three phases of JPEG compression? Discrete cosine transformation, quantization, run-length encoding 2. If a statistical multiplexor is connected to 20 devices, does it require a high speed output line that is equivalent to the sum of the 20 transmission streams? No. Assume that only a percentage of all input streams will have data to transmit at one given time. 3. Run-length encoding can be used to compress what kind(s) of data? It can compress data that cannot be lost during compression and data with a majority of one type of symbol. 4. What are the main differences between statistical time division multiplexing and synchronous time division multiplexing? Sync: continuous bit stream, never out of order; Stat: packets or frames of data, order may vary. 5. What are the two basic forms of compression? Lossless and lossy 6. How are a T-1 and SONET similar? Both are synchronous multiplexing, non-stop, 8000 frames per second 7. How does code-division multiplexing distinguish one signal from another? Each signal is transmitting using a unique binary code or sequence. 8. How many separate channels does a T-1 multiplexor combine into one stream? 24.

6 Fundamentals of Networking and Data Communications, Sixth Edition How is discrete multitone different from the other multiplexing techniques? How is it similar? Same: multiple channels combined over one medium Different: the multiple channels are all intended for one user; each channel is modulated separate from the others 10. How does a synchronous time division multiplexor stay synchronized with the demultiplexor on the receiving end? There are sync bits inserted in the transmission stream at regular intervals. 11. What is the difference between dense wavelength division multiplexing and coarse wavelength division multiplexing? The difference is the number of concurrent lambdas. 12. What would happen if a synchronous time division multiplexor sampled the incoming signals out of order? The demultiplexor would not know what was what. 13. How many different wavelengths can dense wavelength division multiplexing place onto one connection? Potentially 100s 14. In what order does synchronous time division multiplexing sample each of the incoming signals? Round robin order 15. What type of medium is required to support wavelength division multiplexing? Fiber optic cable 16. Frequency division multiplexing is associated with what type of signals? Analog signals 17. Why is addressing of the individual data streams necessary for statistical multiplexing? It is necessary because the order of data streams can vary depending on demand.

7 Fundamentals of Networking and Data Communications, Sixth Edition List three common examples of frequency division multiplexing. Broadcast television, radio, cable television Suggested Solutions to Exercises 1. MP3, JPEG, and MPEG all rely on what characteristic in the data in order to perform compression? The characteristic is that some loss of data will not be noticed by the end user. 2. A discrete multitone system is using a modulation technique on its subchannels, each of which generates a 64-kbps stream. Assuming ideal conditions (no noise), what is the maximum data rate of the discrete multitone system? Assuming the system supports 256 simultaneous channels, 256 x 64 kbps = 16,384,000 bps. 3. Can you compress a set of bank statements using JPEG compression? Explain. You can, but it is better not to, since JPEG is lossy. 4. The telephone company has a fiber-optic line with time division multiplexing that runs from the United States to England and lies on the ocean floor. This fiber-optic line has reached capacity. What alternatives can the telephone company consider to increase capacity? A good possible solution is wavelength division multiplexing. 5. Given the following bit string, show the run-length encoding that would result: , 5, 0, 18, 6, 0, 0, 10 Or 4-bit binary nibbles: 0111, 0101, 0000, 1111, 0011, 0000, 0110, 0000, 0000, When data is transmitted using a statistical multiplexor, the individual units of data must have some form of address that tells the receiver the intended recipient of each piece of data. Instead of assigning absolute addresses to each piece of data, is it possible to incorporate relative addressing? If so, explain its benefits. Yes, relative addressing can be used. You only need to create an address for those devices connected to this pair of multiplexors. This should be a smaller address space than all the devices in the entire system.

8 Fundamentals of Networking and Data Communications, Sixth Edition If data has a large number of one type of symbol, which type of compression would be the most effective? Run-length encoding 8. Ten computer workstations are connected to a synchronous time division multiplexor. Each workstation transmits a 128 kbps. At any point in time, 40 percent of the workstations are not transmitting. What is the minimum necessary speed of the line leaving the multiplexor? Will the answer be different if we use a statistical multiplexor instead? Explain your reasoning. Sync TDM: 10 workstations x 128 kbps = total speed of high-speed line. Doesn t matter if a device is transmitting or not, a space is still allocated for each device. Stat TDM: It should be different, because a stat TDM only reserves a space for devices that currently have data to transmit. So should be 40% capacity of sync TDM. 9. Is the form of DSL that a company uses different from the form of DSL that a home user subscribes to? Explain. More than likely, it is different.. Corporate DSL is typically faster, thus more expensive, and it is often symmetric. 10. What is the synchronization bit in a T-1 frame used for? Why is it necessary? The synchronization bit is necessary to keep the receiver synchronized with the incoming data stream. 11. In theory, code division multiplexing can have 264 different signals in the same area. In reality, this is not possible. Why not? Show an example. What happens if you assign one mobile device the code and a second device the code ? If first device transmits a 1 and second device transmits a 0, they would be sending the same code. Furthermore, the sum of signals would be 0. Is this a positive or negative? 12. If only four computers are transmitting digital data over a T-1 line, what is the maximum possible data rate for each computer? It doesn t matter how many are currently transmitting, T-1 allocates space for 24 separate devices.

9 Fundamentals of Networking and Data Communications, Sixth Edition Which of the multiplexing techniques can be used on both conducted media and wireless media, which on only conducted media, and which on only wireless media? Conducted and wireless: frequency division and time division Only on conducted: wavelength division Only on wireless: code division (can actually be done on conducted media but most if not all examples are performed on wireless) 14. Twenty-four voice signals are to be multiplexed and transmitted over twisted pair. What is the bandwidth required (in bps) if synchronous time division multiplexing is used, if we use the standard analog to digital sampling rate, and if each sample is converted into an 8-bit value? Each voice channel has a bandwidth of 4000 Hz. Sampled two times the frequency equals 8000 samples per second. Each sample is then converted into an 8-bit value, so 8000 samples per second times 8-bits per sample = 64,000 bits per second; 24 signals times 64,000 bps each equals 1,536,000 bps. 15. Why is wavelength division multiplexing more like frequency division multiplexing and less like time division multiplexing? In wavelength division multiplexing different wavelengths of color lasers are used, and different wavelengths of color are simply different frequencies of radio signals. 16. Twenty-four voice signals are to be multiplexed and transmitted over twisted pair. What is the total bandwidth required if frequency division multiplexing is used? Each voice channel has a bandwidth of 4000 Hz. 24 times 4000 Hz equals Hz. 17. How many frames per second does a T-1 and SONET transmit? Why this number? 8000 frames per second; typical voice range of 4000 Hz, sampled twice per highest frequency (Nyquist rule), yields 8000 samples per second 18. A benefit of frequency division multiplexing is that the all the receivers do not have to be at the same location. Explain what this benefit means and give an example. All people watching television do not have to be in the same place. Their de-multiplexors (television sets) can be located anywhere.

10 Fundamentals of Networking and Data Communications, Sixth Edition Mobile user A is using code division multiplexing and has been assigned a binary code of Mobile user B, also using code division multiplexing, has been assigned a binary code of Mobile user A transmits a 1, while Mobile user B transmits a 0. Show the sum of products that results and your calculations. A sends a 1, so code sent is: B sends a 0, so code sent is: Total: 0, -2, 0, -2, 2, 0, 2, 0 A s original signal: -1, -1, -1, -1, 1, 1, 1, 1 Products: 0, 2, 0, 2, 2, 0, 2, 0 Sum of products: 8 B s original signal: -1, 1, -1, 1, -1, 1, -1, 1 Products: 0, -2, 0, -2, -2, 0, -2, 0 Sum of products: Compared to the other multiplexing techniques, state two advantages and two disadvantages of each of the following: a. frequency division multiplexing b. synchronous time division multiplexing c. statistical time division multiplexing d. dense wavelength division multiplexing FDM Adv: simple, popular, all receivers don t have to be in same place. FDM Disadv: analog signals and noise Sync TDM Adv: less noise (digital), can apply digital techniques Sync TDM Disadv: never-ending stream is not an efficient use of medium Stat TDM Adv: less noise (digital), can apply digital techniques, more efficient use of medium Stat TDM Disadv: May not be able to support all inputs at the same time. WDM Adv: incredible capacity WDM Disadv: cost. 21. The cell phone company in town uses code division multiplexing to transmit signals between its cell phones and the cell towers. You are using your cell phone while standing next to someone using her cell phone. How does the system distinguish the two signals? Essentially each cell phone is assigned a different bit or chip sequence. When a cell phone transmits a binary 1, it is using its assigned sequence. The receiver can separate one cellphone s sequence from another. Thinking Outside the Box 1. Wireless? Such as terrestrial microwave or free space optics? What about running a couple strands of fiber and doing stat TDM multiplexing? I would avoid copper-based media.

11 Fundamentals of Networking and Data Communications, Sixth Edition Either wireless as above, or call the telephone company and see what they can offer. Right-ofway is definitely a problem here. 3. The only connection is the need for the receiver to stay in sync with the incoming bit stream. 4. The newer techniques like discrete multitone are getting much more complex by trying to combine separately modulated signals over one medium. 5. You might see a slowdown, as DSL is affected by noise, and AM radio might be generating some of the same frequencies as used by DSL. But most DSL systems now transmit signals that avoid AM signals x 800 x 24 bits/pixel x 30 frames/second x 7200 seconds/2 hours = x 1012 bits for a two hour (7200 seconds) movie. A DVD holds 4.7 GB, or 3.76 x 1010 bits. So a movie has 100 times more data than a DVD.