The Bandwidth of a Communication Medium Is Measured as

Reliable Communication

InfiniBand, iWARP, and RoCE

Manoj Wadekar , in Handbook of Fiber Optic Data Communication (Fourth Edition), 2013

11.4.1.1 Reliable connection and reliable datagram

For reliable communication, data is delivered reliably through a combination of sequence numbers and acknowledgment messages (ACK/NAK). Upon detecting an error or loss of packet, the source can recover by retransmitting the packet without involvement from the user application. This mode guarantees the delivery of a message packet exactly once. When applications need to rely on the underlying transport to guarantee delivery of messages to its destination, this mode is used. This mode frees up an application to safeguard against the unreliability of the underlying media or delivery mechanisms.

Reliable connection (RC) mode provides reliable data transfer between nodes using a direct dedicated connection between the source and destination end nodes.

Reliable datagram (RD) mode provides reliable packet message delivery to any end node without a dedicated connection between the source and destination end nodes. This is an optional mode.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780124016736000118

Fundamentals and Standards of Compression and Communication

Stephen P. Yanek , ... Joan E. Fetter , in Handbook of Medical Imaging, 2000

3.1 Signal Hierarchy and Transfer Rates

To achieve reliable and effective communications, signals must be accurately generated arid propagated. Signals associated with telecommunications systems are usually stated in rather self-descriptive terms, for example, digital signal (DS), optical carrier (OC), or synchronous transport signal (STS). Signals, which contain information, are typically described as frequencies. The bandwidth needed to convey the information is the difference between the highest and lowest frequencies of the signals containing the information. The bandwidth of a communications channel is the difference between the highest and lowest frequencies that the channel can accommodate. The bandwidth of the communications channel must be equal to or greater than the bandwidth of the information-carrying signals. Thus, a communications channel that carries the range of voice frequencies (300 to 3000 Hz) must have a bandwidth of at least 3000 Hz. In contrast, approximately 200 KHz of bandwidth is required for FM transmission of high-fidelity music, and 6 MHz for full-motion, full-color television signals.

A digital communications system uses digital pulses rather than analog signals to encode information. The North American high-speed digital telephone service is referred to as the T carrier system. The T carrier system uses pulse code modulation (PCM) techniques to sample and encode information contained in voice-grade channels, and then time division multiplexing (TDM) techniques to form a DS from a group of 24 voice-grade channels. The information carrying signal of each channel is sampled 8000 times per second. The sample is represented or encoded using 8 bits, thus forming frames of information at a rate of 64 bps. Details about the T carrier system are summarized in Table 4. Metrics are bit transfer rates, the number of voice frequency analog signals that are multiplexed to form a DS, and the number of 6 MHz TV channels that can be transmitted via a T carrier.

TABLE 4. T-carrier baseband system [10]

T carrier designator	Data rate (Mbits/sec)	Digital signal type	Voice grade channels	TV channels	Medium
T1	1.544	DS-1	24	—	Wire pair
T2	6.312	DS-2	96	—	Wire pair
T3	44.736	DS-3	672	1	Coax, fiber
T4	274.176	DS-4	4032	6	Coax, fiber
T5	560.160	DS-5	8064	12	Coax

A single DS-1 signal is usually transmitted over one pair of twisted wires of either a 19-gauge or 22-gauge, known as a T1 line span. Two lines, one for transmit and one for receive, are used in a line span. Repeaters are required about every mile to compensate for power loss. The assemblage of equipment and circuits is known as the T1 carrier system. The lengths of T1 carrier systems range from about 5 miles to 50 miles. The T2 carrier system uses a single 6.312 Mbps DS for transmission up to 500 miles over a low capacitance cable. A T3 carrier moves 672 PCM encoded voice channels over a single metallic cable.

With the development of fiber optic telecommunications systems, signaling standards that were adequate for wire pairs and coaxial cable warranted revision. Fiber-based telecommunications systems are virtually error-free and are capable of reliably moving signals more rapidly than wire systems. The American National Standards Institute (ANSI) published a new standard called Synchronous Optical Network (SONET) in 1988 [10]. It was known as ANSI T1.105 and evolved into an international standard that was adopted by CCITT in 1989. The OC-1 signal is an optical signal that is turned on and off (modulated) by an electrical binary signal that has a signaling rate of 51.84 Mbits/sec, the fundamental line rate from which all other SONET rates are derived. The electrical signal is known as the STS-1 signal (Synchronous Transport Signal — level 1). OC-N signals have data rates of exactly N times the OC-1 rate. Table 5 lists standard transmission bit rates used with SONET and their equivalent STSs.

TABLE 5. SONET signal hierarchy [10]

OC level	Data rate (Mbits/sec)	Synchronous transport signal	Number of DS-Is
OC-1	51.84	STS-1	28
OC-3	155.52	STS-3	84
OC-9	466.56	STS-9	252
OC-12	622.08	STS-12	336
OC-18	933.12	STS-18	504
OC-24	1244.16	STS-24	672
OC-36	1866.24	STS-36	1008
OC-48	2488.32	STS-48	1344

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780120777907500540

Antennas, Diversity, and Link Analysis

Vijay K. Garg , in Wireless Communications & Networking, 2007

10.2 Antenna System

The radio communication system requires a reliable communication between a fixed base station and a mobile station. The goal of the system designer is to have the same performance in both the transmitting and receiving directions. This is not always possible since the base station typically has a higher output power than the mobile station. Also, the mobile station antenna is typically on the street at a height of 1 to 2 m compared to the base station antenna height of 50 to 60 m. This results in the mobile receiver having a high noise level due to interference caused by the ignitions of nearby vehicles or other objects, whereas the base station, with its higher antenna, usually sees a quieter radio environment. These factors can combine to favor one particular direction over the other. The wireless system designer must carefully consider all these factors or the range of the system may be limited by poor performance in one direction.

The types of antennas, their gain and coverage patterns, the available power to drive them, the use of simple or multiple antenna configurations, and the polarization (the polarization of a radio wave is the orientation of its electric field vector) are the major factors that can be controlled by the system designer. The system designer has no control over the topography between the base station and mobile station antennas, the speed and direction of the mobile station, and the location of antenna(s) on the mobile station. Each of these factors significantly affects system performance. Sometimes the placement of the mobile antenna can severely limit system performance. While the mobile antenna is usually installed by a knowledgeable technician, the owner of the vehicle may force a nonoptimum placement of the antenna. Furthermore, cellular antennas are vertically polarized physically, but many vehicle antennas are no longer vertical after the vehicle is sent through a car wash.

Along with the type of antenna there is the relative pattern of the antenna, indicating in which direction the energy emitted or received will be directed. There are two primary classifications of antennas: omnidirectional and sectorized (directional). The omnidirectional antennas are used when the desire is to have a 360° radiation pattern. The sectorized antennas are used when a more refined pattern is needed. The directional pattern is usually required to facilitate system growth through frequency reuse.

The choice of antenna directly impacts the performance of both the base station and the overall network. The antenna selection must consider a number of design issues. Some involve the antenna gain, the antenna pattern, the interface or matching to the transmitter, the receiver used for the site, the bandwidth and frequency range over which the signals will travel, and the power handling capabilities.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780123735805500442

Layer 4: The Transport Layer

In Hack the Stack, 2006

SSL/TLS Summary

The Transport layer is intended to provide reliable communications between two endpoints. SSL and TLS build upon the traditional functionality of TCP to provide confidentiality (by encryption) and integrity (via hashing and digital signatures). SSL is layered between TCP and the applications that use the protocol.

It is no secret that Web browsers and servers are the biggest SSL customers. Most programs that employ SSL support not only the latest versions of the software but are also back-ward compatible with some prior versions. You usually find SSL version 2, SSL version 3, and TLS version 1, actively supported. All things considered, these tools have proven to be very competent, which is witnessed by the millions of secure exchanges sent each day in support of e-commerce and so forth. As a parting thought, you may want to find a copy of stunnel (try www.stunnel.org) to do some experimenting with this technology.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9781597491099500095

Classical Error-Correcting Codes

Richard Hamming , in Classical and Quantum Information, 2012

Coding theory is an application of information theory critical for reliable communication and fault-tolerant information storage and processing; indeed, the Shannon channel coding theorem tells us that we can transmit information on a noisy channel with an arbitrarily low probability of error. A code is designed based on a well-defined set of specifications and protects the information only for the type and number of errors prescribed in its design. A good code has the following characteristics:

1.

It adds a minimum amount of redundancy to the original message.

2.

Efficient encoding and decoding schemes for the code exist; this means that information is easily mapped to and extracted from a codeword.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780123838742000047

Secure Network Coding

Sidharth Jaggi , Michael Langberg , in Network Coding, 2012

4.1 The Coherent Case

For the problem of error correction we first study the rate of reliable communication in the coherent setting in the presence of an active jammer that can jam z_O links of the network and observe all links of the network. In a nutshell, we show that this rate is C − 2z_O for C = 2z_O and 0 otherwise. Namely, the rate is equal to (C − 2z_O )⁺. We start by considering the class of one-hop unicast networks. In a one-hop unicast network there is a single source s that wishes to communicate with a single terminal t over C multiple (s, t) links. We assume that the links may carry a single character from a given alphabet ∑ of size q, and that the source wishes to transmit R characters of ∑ to t. It is not hard to verify that the task of designing a communication scheme with rate R that allows reliable communication over one-hop unicast networks in the presence of an adversary that may jam z_O of the links is equivalent to the design of [C, R] error correcting codes that are resilient to z_O errors (i.e., have minimum distance 2z_O + 1). ¹

Let q be the size of ∑. There are multiple bounds on the rate R(C, z_O , q) of error correcting codes over alphabets of size q with block length C and minimum distance 2z_O + 1. It is well-known that $R\left(C,z_O,q\right)\le C-\mathrm{\Delta }$ for $\mathrm{\Delta }={\log }_q\left({\sum }_{i=0}^{z_O}\left(\genfrac{}{}{0ex}{}{C}{i}\right){\left(q-1\right)}^i\right)$ , e.g., [48]. This bound is referred to as the sphere packing or Hamming bound, and follows from a simple volume argument. As q approaches infinity it can be verified that this bound approaches C − z_O . This bound holds for all types of errors—random or adversarial. Further, the Singleton bound (e.g., [48]), derived using the pigeonhole principle, shows that $R\left(C,z_O,q\right)\le C-2z_O$ . Hence for sufficiently large alphabet sizes q, the Singleton bound is tighter than the sphere packing bound.

What about lower bounds on R(C, z_O , q)? Several coding techniques [48] (including for example Read-Solomon codes) imply C-block error correcting codes resilient to z_O errors whose rate equals C − 2z_O . Most relevant to this chapter are the works of Gilbert [18] and Varshamov [79] that show that $R\left(C,z_O,q\right)\ge C-\mathrm{\Delta }$ where $\mathrm{\Delta }={\log }_q\left({\sum }_{i=0}^{2z_O}\left(\genfrac{}{}{0ex}{}{C}{i}\right){\left(q-1\right)}^i\right)$ . Notice that the summation in this case is from 0 to 2z_O (as apposed to z_O in the Hamming bound). The discussion above implies that as q tends to infinity, the Singleton bound is tight and corresponds to the capacity of one-hop unicast networks in the presence of jammers.

A natural and intriguing question is whether the above setting also holds in more complicated networks as well. This question was studied by Cai and Yeung in [10, 87] and was answered in the affirmative. Namely, [10, 87] show an analog to the Hamming bound, Singleton bound, and Gilbert-Varshamov bound in the coherent network coding setting. Moreover, they show their Singleton-type bound for networks equals their Gilbert-Varshamov type bound for large values of q. ² The crux of their analysis lies in understanding the combinatorial nature of information transmitted on minimum cut-sets of the network that separate source terminal pairs. ³ In what follows we give an overview of the results in [10, 87].

Consider any given network G = (V, E) with (error free) capacity C. Let A and B be a partition of V, and let cut(A, B) denote the set of links directed from a node in A to a node in B. To obtain an analog to the Hamming bound for networks, [87] considers the information transmitted over cut-sets (A, B), or to be precise, the mapping between the source information X and the information Z^m = Z ₁,…,Z_m transmitted over the cut-set. Here $m=\mathfrak{|}\text{\hspace{0.17em}cut}\left(A,B\right)\mathfrak{|}$ . Roughly speaking, if there are no links directed from B to A in G, it must be the case that Z^m is an [m, 2z_O + 1] error correcting code. This follows directly by the fact that decoding at terminal t is solely a function of Z^m . Indeed, if Z^m did not have minimum distance 2z_O + 1 then a malicious jammer corrupting z_O links from cut(A, B) may cause a decoding error at t. Note that the reduction above relies on the lack of edges from B to A, otherwise errors on certain links of cut(A, B) may affect other links in cut(A, B) (such effects do not occur in the standard model of error correcting codes). Once the reduction between network communication and error correcting codes is established, the Hamming-type bound and Singleton-type bound follow.

Theorem 4 Network Hamming Bound

Let G be an acyclic network with (error free) cut capacity C, in which each link can carry a single character of an alphabet ∑ of size q. Then the coherent capacity when at most z_O of the links of G are jammed is at most $C-\mathrm{\Delta }$ where:

$\mathrm{\Delta }={\log }_q\left(\underset{i=0}{\overset{z_O}{{{\displaystyle \sum }}^{\text{}}}}\left(\begin{array}{c}C\\ i\end{array}\right){(q-1)}^i\right).$

As the field size q approaches ∞ with fixed C and z_O, this bound approaches C − z_O.

As for classical error correcting codes, a stronger bound for the network adversarial error case for large q is the network analog of the Singleton bound.

Theorem 5 Network Singleton Bound

Let G be an acyclic network with (error-free) cut capacity C. Then, the coherent capacity in the presence of an adversary that may jam up to z_O of the links of G is at most (C − 2z_O)⁺.

We now turn to discuss lower bounds on the coherent capacity in the presence of an adversary that may jam up to z_O links. In [10] a Gilbert-Varshamov bound in the context of network communication is derived. It is well-known that in the error-free coherent setting, one can communicate the set ∑ ^C of distinct messages successfully over the network using, for example, linear network codes that are constructed at random. Using such network codes, the main idea in [10] is to carefully construct a subset of messages W ⊂ ∑ ^C with the property that no matter which error pattern is chosen by the adversary, each terminal is able to correctly distinguish the message w ∈ W transmitted. Namely, two words x and x′ of ∑ ^C are said to be non-separable if there exist two error patterns e and e′ such that the information reaching a terminal node when x is transmitted and the adversary applies the error pattern e is identical to that received when x′ is transmitted and e′ applied. The objective in [10] involves identifying a large subset W for which each $w\ne w^{\mathfrak{\prime }}\mathfrak{\in }W$ are separable. The crux of their analysis lies in a careful study, for a given x ∈ ∑ ^C , of the subset V_x of possible words x′ such that x and x′ are non-separable. Bounding the size V of V_x and following the greedy technique of Gilbert [18] will yield sets W of size q^C /V. Moreover, using a Varshamov-type approach one is able to bound V by q ^2z_O and obtain a linear W of size q ^{C − 2z_O} .

Theorem 6 Network Gilbert-Varshamov Bound

Let G be an acyclic network with (error free) cut capacity C in which each link can carry a single character of an alphabet ∑ of size q. If q is sufficiently large, then the coherent capacity in the presence of an adversary that may jam up to z_O of the links of G is at least (C − 2 z_O)⁺.

Corollary 1 Coherent Capacity

Let G be an acyclic network with (error free) cut capacity C. Then, the coherent capacity in the presence of an adversary that may jam up to z_O of the links of G is (C − 2z_O)⁺.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B978012380918600007X

Networking, Security, and the Firewall

Brad Woodberg , ... Ralph Bonnell , in Configuring Juniper Networks NetScreen & SSG Firewalls, 2007

Layer 4: The Transport Layer

The transport layer provides a total end-to-end solution for reliable communications. TCP/IP relies on the transport layer to effectively control communications between two hosts. When an IP communication session must begin or end, the transport layer is used to build this connection. The elements of the transport layer and how it functions within TCP/IP are discussed in more detail later in the chapter. The transport layer is the layer at which TCP/IP ports listen. For instance, the standard port which HTTP listens on is TCP Port 80, although HTTP could really run on any TCP port; this is the standard. Again, there is no difference between TCP port 80, 1000, or 50000; any protocol can run on it. Standardized port numbers are used to help ease the need to negotiate the port number for well-known applications.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9781597491181500034

Networking Embedded Systems

Edward Insam PhD, BSc , in TCP/IP Embedded Internet Applications, 2003

Why Network Embedded Systems?

Networking is the common term given to the methods and techniques used for providing reliable communications between geographically remote devices. Networking requires the enforcement of common standards and the use of compatible hardware and software environments. This guarantees that communications can take place between machines that may be completely different in architecture and operating systems. It is understandable that a lot of design effort may need to be put in place to ensure that systems are compatible and can talk to each other without problems.

Figure 1-1 shows two applications where networked embedded systems may be used. Figure 1-1(a) shows a point of sale vending machine with an internal modem linked to head office via a phone line. The processor within the vending machine may be a standard microcontroller with its normal peripherals: RAM, ROM and I/O ports used for driving the various lamps, solenoids and push-button switches. The auto-dial modem is used to connect to the company's host over the normal telephone network. The central host computer may accept calls or interrogate the remote, perhaps by dialing its number at regular hourly or daily intervals. The vending machine could also originate calls by dialing head office directly; these calls could be local or long distance depending on location. There have to be enough modems working in parallel at head office to ensure incoming calls are not lost, and avoid engaged tones. With several thousand remotes in a large geographical area, this could result in major investment in plant and equipment, not including manpower, maintenance, and the cost of all the long-distance phone calls.

Figure 1-1. Examples of networked embedded systems

Fortunately, the company could make use of the Internet. Using the Internet, the remotes only need to make local rate telephone calls to an Internet Service Provider (ISP), very much in the same way a domestic user dials their local ISP to use the Internet. Head office only needs to have a direct dedicated connection to the Internet. No multiple phone lines, no modems, no long distance telephone calls. Savings are not only made in telephone system infrastructure, but also in maintenance and labour required to support in-house equipment and plant. The ISP will provide all these facilities at competitive costs.

Figure 1-1(b) shows another application where a CCTV camera uses the existing local area computer network within a building to send security images from a parking lot. Images are digitized within the camera and sent in 'network friendly' encoded form to central office, which decompresses the images, and displays them on a screen. The embedded processor within the camera needs to be fast enough to capture, process and encode the images more or less in real time. This requires very fast CPU architectures similar to those found inside PCs. Alternatively, the camera could make use of dedicated integrated circuits or custom gate arrays. The diagram shows the camera using a wireless network. These use radio waves instead of cabling, and could be of benefit in areas where it is difficult to wire cables all the way to the camera.

Both these examples are typical of embedded systems networking. By sharing the use of an existing networking infrastructure (Internet or local area network) big savings can be made.

Adding networking to an embedded system is not a trivial task. From the relatively safe confines of an appliance, networking involves sending data over a hostile medium (the world outside), and communicating with a hostile remote system, which may not really care what status our device is in, and which may be operating under different operating systems, clocks or time frames. It soon becomes obvious to the designer that serious amount of resources will have to be allocated for the proper design and implementation of good, reliable networking. This includes not just the software, but also the hardware required for providing electrical isolation, decoupling and mechanical stability. By definition, embedded systems are resource limited systems (resources in this context, are the components of a system including ROM and RAM space, I/O ports and even CPU time), and an embedded system with surplus resources is an over-engineered system. Adding networking here may mean a rethink and redesign of the same product with more components, more memory, more interfaces and more software.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B978075065735850028X

Automotive and Aerospace Systems

Marilyn Wolf , in Computers as Components (Fourth Edition), 2017

9.3.2 Other automotive networks

Time-triggered architectures

The time-triggered architecture [Kop03] is an architecture for networked control systems that provides more reliable communication delays. Events on the time-triggered architecture are organized around real time. Since devices in the network need time to respond to communication events, time is modeled as a sparse system. Intervals of active communication are interspersed with idle periods. This model ensures that even if the clock's value varies somewhat from device to device, all the devices on the network will be able to maintain the order of events in the system.

FlexRay

The FlexRay network [Nat09] has been designed as the next generation of system buses for cars. FlexRay provides high data rates—up to 10   Mbits/s—with deterministic communication. It is also designed to be fault tolerant. Communications on the bus are designed around a communication cycle. Part of the cycle, known as the static segment, is dedicated to events for which communication time has been guaranteed. The timing of communication for each of these events determined by a schedule set up by the designer. Some devices may need only sporadic communication, and they can make use of the dynamic segment for these events.

LIN

The Local Interconnect Network (LIN) bus [Bos07] was created to connect components in a small area, such as a single door. The physical medium is a single wire that provides data rates of up to 20   kbits/s for up to 16 bus subscribers. All transactions are initiated by the master and responded to by a frame. The software for the network is often generated from a LIN description file that describes the network subscribers, the signals to be generated, and the frames.

Several buses have come into use for passenger entertainment. Bluetooth is becoming the standard mechanism for cars to interact with consumer electronics devices such as audio players or phones. The Media Oriented Systems Transport (MOST) bus [Bos07] was designed for entertainment and multimedia information. The basic MOST bus runs at 24.8   Mbits/s and is known as MOST 25, and 50 and 150   Mbits/s versions have been developed. MOST can support up to 64 devices. The network is organized as a ring.

Data transmission is divided into channels. A control channel transfers control and system management data. Synchronous channels are used to transmit multimedia data; MOST 25 provides up to 15 audio channels. An asynchronous channel provides high data rates but without the quality-of-service guarantees of the synchronous channels.

The next example looks at a controller designed to interconnect LIN and CAN buses.

Example 9.3 Automotive Central Body Controllers

A central body controller [Tex11C] runs a variety of devices that are part of the car body: lights, locks, windows, etc. It includes a CPU that performs management, communications, and power management functions. The processor interfaces with CAN bus and LIN bus transcenvers. Devices such as remote car locks, lights, or wiper blades are connected to LIN buses. The processor transfers commands and data between the main CAN bus and the device-centric LIN buses as appropriate.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780128053874000091

Wireless Communications

Michele Zorzi , A. Chockalingam , in Encyclopedia of Physical Science and Technology (Third Edition), 2003

II.B.2 The IS-95 CDMA System

As already mentioned, IS-95 is based on CDMA, which takes advantage of spread-spectrum modulation in order to guarantee reliable communication in the presence of interference and multipath. spread-spectrum communication technique has its origins in military communications where antijamming (AJ) and low probability of intercept (LPI) capabilities are the main objectives. The information bearing signal (voice or low speed data), which is typically narrowband in nature (few tens of KHz bandwidth), is spread over a much wider bandwidth (typically of the order of several MHz) before transmission. The ratio of the transmission bandwidth (spread bandwidth) to the information bandwidth is known as the processing gain. The larger the processing gain, the better the system's immunity to jamming and intercept. Typical processing gains in the range 100 to 1000 are common.

Spectrum spreading is typically realized through one of two ways, namely, direct sequence (DS) or frequency hopped (FH) spread-spectrum (SS) technique. In DS-SS, the information signal is multiplied by a spreading signal which is generated at a much faster rate compared to the information rate. In FH-SS, narrowband modulation is used, but the instantaneous frequency of the modulating carrier is varied in a predetermined pattern. The spreading signal in DS and the hopping pattern in FH are generated using pseudo-random noise (PN) sequence generators. In order to correctly demodulate the information signal at the receiver, knowledge of the PN sequence used for spreading at the transmitter is needed. The receiver essentially correlates the received signal with a local replica of the PN signal (the operation is known as despreading), following which the data is demodulated.

Another important application of SS is multiple access, i.e., shared use of a common channel by several transmitters. Different transmitters can be assigned different spreading PN sequences (or spreading codes) and the central station can employ several correlation receivers each tuned to a given transmitter's spreading code. Such a multiple access system which uses DS-SS is known as the direct sequence code division multiple access (DS-CDMA). The IS-95 CDMA system has successfully exploited the multiple access capability of DS spread spectrum in cellular communications.

Since the users are separated in the code domain, CDMA users can transmit simultaneously in time over the entire available bandwidth. Also, due to the fact that the bandwidth of the transmitted signal is much larger than the bandwidth of the information signal (a factor of 128 is specified in IS-95), CDMA can enjoy the benefit of frequency diversity, since different spectral components can fade independently. In the time domain, this corresponds to being able to discriminate among the various replicas of the signal arriving at the receiver with different delays, which can therefore be constructively added rather than randomly interfere. The particular structure which is able to combine signals in this way is called RAKE receiver.

Another advantage of SS systems is the possibility of trading bandwidth expansion for coding redundancy. For example, if the channel bandwidth is 100 times larger than the user's signal bandwidth, it is possible to transmit such signal uncoded with a spreading factor of 100, or coded with a rate-1/2 code and a spreading factor of 50. This allows the use of coding (which is essential in certain applications) without further increase in the bandwidth of the transmitted signal. On the other hand, in TDMA-based systems the use of coding requires bandwidth expansion, unless trellis coded modulation (TCM) is used.

Due to the fact that all users share the same bandwidth and that interferers are essentially seen by the intended receiver as additive noise, any technique that reduces interference or enhances signal detection (i.e., reduces the E _b /N _o required to achieve a desired bit error rate) can result in an improvement of the CDMA system capacity. Also, this characteristic provides graceful performance degradation as users join the system, so that CDMA systems have "soft capacity," in the sense that a user can always be admitted to the system if a slight performance degradation for all users is accepted. This is not possible in static resource assignment systems (such as TDMA and FDMA with fixed channel allocation), where there is a constant number of physical channels per cell, and one more cannot be added. However, dynamic resource assignment has the potential to offer soft capacity even in these systems.

Typical techniques to enhance detection are the use of antenna diversity and of error correction coding (convolutional codes with rate 1/2 on the base-to-mobile link and rate 1/3 on the mobile-to-base link are specified in IS-95). Interleaving is also used to improve the error correction coding performance. Typical techniques to reduce interference are cell sectorization (120° antennas are used, which pick up only one-third as much interference as omnidirectional antennas) and silence suppression (during periods when the user is silent, which amount to more than half the time in typical telephone conversations, the transmitter switches to a low-rate mode used for signaling and synchronization purposes only).

A key requirement for DS-CDMA systems to work properly is that all signals reach the receiver with the same power, as power imbalances may greatly reduce the system capacity. To this purpose, tight power control must be used in the mobile-to-base direction, in order to compensate for the channel attenuation. If perfect power control were possible, all signals coming from mobiles in the same cell would have exactly the same power at the BS. On the other hand, power control errors do occur and interference from other cells cannot be controlled, and these effects degrade the performance of the system. In any event, accurate power control is a critical requirement of CDMA-based systems for proper operation, unlike in other systems where power control may increase the performance to some extent but is by no means critical. Open-loop power control is implemented based on the measurements of the pilot signal transmitted by the BS. Closed-loop power control is also specified in IS-95, with the BS sending one-bit power control commands every 1.25   ms, indicating whether the mobile should turn up or down the transmitted power in steps of 1   dB. A power dynamic range of 80   dB is typically required to track and compensate the received power variations due to fast multipath fading. In the other direction (base-to-mobile), a simpler power control scheme is used, whose main goals are improved transmission quality and minimized interference.

As in other systems, paging and random access channels are present. The broadcast channels are used by the BS to transmit a pilot signal, which is used by the mobiles as a power control reference and also as a coherent reference to be used for signal detection. A synchronization channel is also provided, which is used in the acquisition phase as a timing reference.

MAHO is specified in IS-95, where a mobile monitors the pilot signals coming from a number of cells and reports the results of these measurements to the controller. A unique feature of IS-95 is the so-called "soft handoff," where during the handoff procedure a mobile may be connected to more than one BS simultaneously. This allows the use of macrodiversity reception (the network can combine signals received by multiple BS), and provides a means to avoid interruptions as, unlike in other systems, the connection to the new BS is set up before the connection with the old BS is torn down (i.e., "make before break").

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B0122274105008279

The Bandwidth of a Communication Medium Is Measured as

Source: https://www.sciencedirect.com/topics/computer-science/reliable-communication

Share this post