In Wireless Networks, What Type Of Media Access Control Is Used?

Media Access Control Protocol

Local Area Networks

Judy Wynekoop , in Encyclopedia of Data Systems, 2003

III.B. Media Access

Media access control (MAC) protocols enforce a methodology to permit multiple devices admission to a shared media network. Before LANs, communication between computing devices had been signal-to-signal. That is, two devices were connected by a dedicated aqueduct. LANs are shared media networks, in which all devices attached to the network receive each transmission and must recognize which frames they should have. Media sharing reduced the cost of the network, but also meant that MAG protocols were needed to coordinate use of the medium. There are 2 approaches to media access command in LANs: contention and token-passing.

Contention is a start-come up, first-serve approach. Carrier sense multiple access/standoff detection (CSMA/CD) is the about used contention-based MAC protocol, used in Ethernet networks. When a device must transmit, the NIC monitors the network to make up one's mind whether or not another device is transmitting. The NIC cannot transmit if information technology senses electric signals on the network that indicate some other device is transmitting. Due to propagation delay (the fourth dimension it takes a signal to reach a signal from its sender), it is possible for ii stations to transmit simultaneously or nearly simultaneously, causing a standoff and garbling the letters. When the sending NIC senses that the bulletin propagating along the network is non identical to that which it is transmitting, transmission ceases. The sending NICs and then wait a random amount of time before resending the data. The busier a network is, the more often collisions occur and the longer it takes to transmit data.

CSMA/GA (Collision Avoidance) is a contention method designed to prevent collisions. In CSMA/CA, the NIC monitors the line for a longer time and the line must exist idle for a specified period before a station can transmit. A brusk handshaking packet is then sent before the bulletin. If there is a collision due to simultaneous transmissions, simply the handshaking packets collide and the sending stations await and try once again. CSMA/CA is used in wireless networks, since devices are farther autonomously in a wireless network than in a wired LAN.

Token-passing ensures a sending station has the network'due south entire bandwidth to itself by requiring that a sending station possesses a specific data frame (the token) earlier it tin transmit. On shared-media LANS with high information traffic, greater throughput is achieved using token-passing than a contention-based access method, since there are no collisions with token-passing. Since token-passing is a deterministic admission method (the timing of signals can be predicted), information technology is well suited to time-dependent traffic, such as telemetry. Contention is nondeterministic.

Read full affiliate

URL:

https://www.sciencedirect.com/science/commodity/pii/B0122272404001064

Continued Computing Environment

Debraj De , ... Song Tan , in Advances in Computers, 2013

four.2 Preamble Based MAC Protocol

Asynchronous MAC protocol B-MAC utilizes a long preamble message to achieve low ability advice. But the kind of LPL approaches with long preamble have some drawbacks: (a) excess latency at each hop, (b) suboptimal energy consumption, and (c) suffering from excess energy consumption at non-target receivers. X-MAC [5] proposes solutions to each of these drawbacks by employing a shortened preamble arroyo that also keeps the bones advantages of LPL, such as depression power communication, simplicity, decoupling of transmitter and receiver sleep schedules. A visual timeline representation of asynchronous LPL with long preamble and 10-MAC with brusk preamble are shown in Fig. 10. The X-MAC protocol has been implemented in various operating systems for sensor networks such as TinyOS, Mantis Operating Organisation (MOS), etc.

X-MAC reduces the overhearing problem due to long preamble past dividing that i long preamble into a number of short preamble packets, each containing the ID of the destination node. The stream of short preamble packets finer constitutes a single long preamble. When a node wakes upward and receives a curt preamble packet, it compares the destination node ID (included in the parcel) with it's own ID. If the node is not the destination, information technology returns to sleep way immediately. But if the node itself is the intended destination, it remains awake for the subsequent data bundle.

In improver to shortening the preamble, Ten-MAC also addresses the problem of multiple transmitters sending the entire preamble even though the receiver is already awake. In such situation for X-MAC, when a transmitter is attempting to send but detects a preamble and is waiting for a clear channel, the node listens to the channel and if it hears an acknowledgment frame from the node that it wishes to send to, the transmitter will backoff a random amount and and then send its information without a preamble. The randomized backoff is necessary considering at that place may exist more than one transmitter waiting to send, and the random backoff volition mitigate collisions betwixt multiple transmitters.

In conclusion, in this department we take presented applied arrangement solutions to the challenges in link layer of Wireless Sensor Networks. This includes the works in: link layer compages blueprint and MAC protocols.

Read full chapter

URL:

https://www.sciencedirect.com/scientific discipline/article/pii/B9780124080911000014

Communication Network Compages

Vijay K. Garg , Yih-Chen Wang , in The Electric Applied science Handbook, 2005

6.2.8 ANSI X3T9.5—Fiber Data Distribution Interface

The other type of LAN is called metropolitan surface area network (MAN). MAN has characteristics of high capacity of supporting at least 100-Mbps speed and more than 500 stations on the network. It has a larger geographic scope than LAN provides back up for integrated information types, and has the provision of dual cables to meliorate throughput and reliability

Fiber data distribution interface (FDDI) is an case of Human being that is an ANSI X3T9.5 standard, proposed past the ANSI XT9.5 study group. Each field in the MAC frame can either be represented equally a symbol (four $.25, for nondata) or in bits format (data or address). The IEEE 802.5 token ring frame format is similar to the FDDI frame format. The FDDI frame includes a preamble to help in clocking because this is desirable for high-speed communication. FDDI does not include an access control (AC) field that is used for priority reservation scheme. FDDI uses a capacity allocation scheme instead of priority reservation. Effigy six.8 shows the frame formats of IEEE 802.5 and FDDI.

FDDI MAC protocol is very similar to IEEE 802.5–Token Passing Ring protocol, simply there are differences described as follows:

•: After absorbing the free token, an FDDI station starts to transmit data frames. In IEEE 802.5—Token Passing Ring MAC, the token type chip is inverse to a busy token type, and the data is appended to the token.
•: An FDDI station frees the token right after manual of a data frame, and it will not expect for the data frame return. In IEEE 802.5, a station will not gratis a token until the leading edge of the data frame returns.
•: FDDI MAC uses the Time-Token Protocol (TTP) for both synchronous and asynchronous services to all stations. If a token arrives earlier based on the rotation timer, a station tin optionally send the asynchronous data. If the token is tardily, only the synchronous data can be sent. In IEEE 802.5, MAC protocol is based on explicit priority reservation. Both protocols permit the network to answer to changes in traffic load, but FDDI supports more steady load because lower-priority traffic may take more opportunity to send when a token arrives early on.
•: The use of a restricted token volition allow for two stations to have multiframe dialog capability to interchange long sequences of data frames and acknowledgments. This would improve the functioning of the application that uses the capability.

The data-encoding scheme used past FDDI is called 4B/5B, which encodes 4-bit data inside a 5-bit cell. There is no more than three consecutive zero bits in a prison cell, and at least two transitions occur in a five-chip cell. The binary flake values are represented with not return to zippo inverted (NRZI), where the transition at the kickoff of the scrap time denotes a binary 1 for that fleck fourth dimension, and no transition indicates a binary 0. Only sixteen out of 32 code patterns are used for data, and other patterns are used for control symbols. Timing jitter is one of transmission impairments in the data communication. The deviation of clock recovery occurs when a receiver attempts to recover clocking every bit well as data. Due to the high speed of transmission, the deviation of the clock is more than severe for FDDI than for IEEE 802.5. The centralized clocking used by the IEEE 802.v network is inappropriate for 100 Mbps, and it requires a complicated and expensive phase lock loop circuitry. Distributed clocking is therefore used by FDDI. With distributed clocking, each station recovers a clock from its incoming signal and transmits out at station clock speed. Each station also maintains its ain rubberband buffer, different the IEEE 802.five network that only designates 1 station to have the elastic buffer.

FDDI specifies three reliability requirements of providing an automated bypass for a bad or ability-off station, using dual rings for easy reconfiguration when one ring is broken, and installing a wiring concentrator. Table 6.four summarizes the comparison betwixt IEEE 802.5 Token Passing Ring and ANSI X3T9.5 FDDI.

TABLE 6.4. Comparing IEEE 802.5 and X3T9.v FDDI

ANSI X3T9.five FDDI	IEEE 802.5 Token Passing Ring
Fiber or twisted pair as transmission medium	Twisted pair or optical fiber as manual medium
100 Mbps	four, 16, or 100 Mbps
Reliability specification	No reliability specification
Maximum of 1000 stations	Maximum of 250 stations
4B/5B encoding	Differential Manchester
Distributed clocking	Centralized clocking
Maximum of 4500 octets of frame sizes	Maximum of 4550 octets for a 4-Mbps network and 18200 octets for 16-Mbps and 100-Mbps networks
Time token rotation	Priority reservation
Token release right after the transmission	Token release after decorated token comes back

FDDI is more than reliable than other local network systems, only it is discouraged due to its high cost. There is a tendency to migrate FDDI into switched courage fast ethernet for the following boosted reasons:

•

More bandwidth to the desktops would require increasing the backbone capacity

•

Because of continued evolution of ethernet compages and its speed, it would make business and technical sense to migrate FDDI backbone into switched fast ethernet. Tabular array 6.5 shows the comparison of FDDI and fast ethernet.

Tabular array half-dozen.v. Comparing FDDI and Fast Ethernet

	FDDI	Fast ethernet
Reliability	Cocky-healing dual rings	Can provide ane fill-in connectedness
Maximum frame size	4500 octets	1518 octets
Performance	Sustained performance with increasing number of stations	Whole network endemic by each user with the switched Ethernet
Distance	Upwardly to 32 Km with fiber	Up to 32 Km with fiber
Price	Close to $1000/port	$100-$150/port

Read full affiliate

URL:

https://www.sciencedirect.com/science/commodity/pii/B9780121709600500748

The Influence of Inter-Domain Mobility on Message Stream Response Time in Wired/Wireless Profibus-Based Networks

Luís Ferreira , Eduardo Tovar , in Fieldbus Systems and Their Applications 2005, 2006

2.1 Basics of the PROFIBUS protocol

The PROFIBUS Medium Admission Control (MAC) protocol uses a token passing process to grant bus access to masters. After receiving the token, a PROFIBUS master is capable of processing transactions during its token belongings fourth dimension (T_TH ), which, for each token visit, is the value corresponding to the difference, if positive, betwixt the target token rotation time (T_TR ) parameter and the real token rotation time (T _RR ). For further details, the reader is referred to [v].

A transaction (or bulletin bicycle) consists on the request or send/asking frame from a primary (the initiator) and of the associated acknowledgement or response frame from a master/slave station (the responder). The response must arrive to the principal before the expiration of the Slot Time (T_SL ), a master parameter.

In order to maintain the logical ring, PROFIBUS provides a decentralized ring maintenance mechanism. Each PROFIBUS master maintains two tables – the Gap List (GAPL) and the List of Active Stations (LAS), and may optionally maintain a Live List (LL).

The GAPL consists of the accost range from 'This Station' accost until 'Next Station' accost, i.due east., the next principal in the logical token ring. Every time the Gap Update Timer (T_GUD ) expires in a master, information technology starts checking the addresses in its GAPL. This is accomplished by inquiring (at most) one master on the GAPL per token visit. If a new principal replies, and so the requesting master passes the token to this new master and Updates its 'Next Station' address. Otherwise, the requesting master continues its operation. In the MLR arroyo, this mechanism is used for enabling the mobility of wireless chief stations, equally detailed later on.

The LAS is a list of all the masters in the logical ring, and the LL contains all active stations (both masters and slaves).

Read total affiliate

URL:

https://www.sciencedirect.com/science/article/pii/B9780080453644500599

Ad Hoc Wireless Sensor Networks (WSNs)

Anurag Kumar , ... Joy Kuri , in Wireless Networking, 2008

10.6.ii IEEE 802.15.4 (Zigbee)

The other sensor MAC protocol that has received wide attention is the IEEE 802.fifteen.4 MAC. The protocol was introduced outset in the context of Depression-Rate Wireless Personal Surface area Networks (LR-WPANs). The PHY and MAC layers in LR-WPANs are defined past the IEEE 802.15.4 group, whereas the higher layers are defined by the Zigbee brotherhood.

IEEE 802.15.4 defines two types of devices: a Full Office Device (FFD) and a Reduced Function Device (RFD). The FFDs are capable of playing the role of a network coordinator, but RFDs are not. FFDs can talk to any other device, while RFDs tin can only talk to an FFD. Thus, one manner of operation of the IEEE 802.fifteen.4 MAC is based on a hierarchy of nodes, with 1 FFD and several RFDs connected in a star topology (see Figure ten.12). The FFD at the hub, which is a network coordinator, plays the function of a cluster-head, and all communication is controlled by information technology. In the peer-to-peer topology, however, all nodes are as capable; all are FFDs.

Figure 10.13 shows the superframe structure defined for IEEE 802.xv.4. The superframe begins with a buoy. Nodes hearing the beacon can set their local clocks accordingly, then that they become to slumber and wake upwardly at the same time. This means synchronized operation.

The superframe is divided into an active and an inactive period. During the inactive period, nodes slumber. The active period consists of at most 3 parts—buoy transmission interval, the Contention Access Flow (CAP) and an optional Contention Gratis Menstruation (CFP). During the CAP, nodes contend using slotted CSMA/CA, as in IEEE 802.11 (see Affiliate vii). In the CFP, a node can be allotted Guaranteed Fourth dimension Slots (GTSs) by the network coordinator. Nodes asking for GTS allocation by sending explicit GTS allocation requests. Transmitted frames are always followed past Inter-Frame Spacings.

Read full chapter

URL:

https://www.sciencedirect.com/science/commodity/pii/B9780123742544500119

Guidelines and criteria for selecting the optimal low-power wide-surface area network technology

Guillermo del Campo , ... Asuncion Santamaria , in LPWAN Technologies for IoT and M2M Applications, 2020

13.ii.2.1 MAC protocol

The use of traditional cellular medium admission control (MAC) protocols that utilize time and frequency diversity, requires precise synchronization, and are incompatible with the low-cost LPWAN terminate-devices. One of the most adopted MAC protocols for LPWAN is the carrier-sense multiple admission with collision avoidance (CSMA/CA), widely implemented in WLAN and LoWPAN technologies. All the same, CSMA/CA becomes less effective when the number of nodes increases (an intrinsic feature of LPWAN). The use of virtual carrier sensing solves this problem, merely it does non behave well with massive deployments. Alternatives to CSMA/CA are the use of ALOHA, a carrier-sensing-less random access protocol for simple and depression-cost transceivers; or TDMA/orthogonal FDMA (OFDMA)-based protocols, resulting in more complex and expensive end-devices.

LoRa (TDMA), SigFox (FDMA), Telensa (FDMA), and Wi-SUN (PCA—pure commonage ALOHA) use the ALOHA protocol. CSMA/CA is implemented past Snowfall (too DOFDM), DASH7, and Wi-Sun. NB-IoT and LTE-CatM employ OFDMA for the downlink communication and unmarried-carrier FDMA (SC-FDMA) for the uplink. GSM-IoT and Weightless apply FDMA and TDMA. Finally, Ingenu-RPMA utilizes code sectionalisation multiple access (CDMA) and MIOTY employs telegram splitting multiple access (TSMA)

Read full chapter

URL:

https://www.sciencedirect.com/scientific discipline/article/pii/B9780128188804000144

Wireless medical sensor networks for smart east-healthcare

Abhinav Adarsh , Basant Kumar , in Intelligent Data Security Solutions for e-Health Applications, 2020

six.3 Hybrid MAC protocols

Shaswat Pathak [52] discussed different MAC protocols used in the BAN part of wireless medical sensor networks and proposed an energy efficient node priority-based MAC protocol for BANs. IEEE 802.fifteen.4, or ZigBee, is a hybrid MAC protocol that aims to utilise the advantages of both these classes of MAC protocol by incorporating a contention- too equally schedule-based structure. It has a contention-based slotted CSMA/CA interval in which nodes contend for medium access and transmit a maximum of seven guaranteed time slots for the emergency data, which is allotted on a schedule-based TDMA scheme. However, this protocol lacks the prioritized access given to nodes that sometimes carry important medical information. For example, a cardiac patient in a critical intendance unit requires the continuous monitoring of ECG rather than temperature and claret pressure data. Assuming node capability to raise an alarm in case of emergency, a prioritized access of medium is required past these nodes. Abhiav Adarsh et al. [63] proposed a MAC protocol that incorporates a data sensitivity-based priority mechanism.

DSA-MAC [63]: This paper presents a data-sensitive adaptive medium access command (DSA-MAC) protocol for an intrahospital scenario. The primary feature of this protocol is to prioritize the node, based on sensitivity of the information it has to communicate. The standard protocol such equally IEEE 802.fifteen.iv for BANs cannot fulfill all the requirements as it does not help or maintain prioritized and varied information transfer requirements of various medical sensor nodes. This paper proposes a DSA-MAC protocol past prioritizing the bachelor medical sensor information for aqueduct allocation. Functioning assay of DSA-MAC shows that it outperforms other conventional MAC protocols in terms of throughput, data standoff ratio, free energy consumption, and average transmission time.

Read full chapter

URL:

https://www.sciencedirect.com/science/commodity/pii/B9780128195116000157

Packet-Switched Networks

Jean Walrand , Pravin Varaiya , in High-Performance Communication Networks (Second Edition), 2000

3.iv FDDI

The Cobweb Distributed Data Interface (FDDI) (see Figure 3.22) is an ANSI (American National Standards Institute) standard for a 100-Mbps network, published in 1987. Until recently, FDDI was the preferred applied science for connecting LANs. Gigabit Ethernet and ATM switches, which operate at higher speeds, are expected to readapt FDDI.

FDDI connects up to 500 nodes with optical fibers, in a dual ring topology. The altitude betwixt adjacent nodes cannot exceed 2 km when multimode fibers and LEDs are used. Longer separation is possible with unmarried-manner fibers and laser diodes. The maximum length of the fibers is 200 km. Because of this length, FDDI networks are used to interconnect computers inside a campus. Many vendors supply FDDI hardware and software for workstations.

The figure shows FDDI networks that connect workstations to file servers and printers, or workstations together, or terminals and terminal emulators to workstations.

As indicated in Figure three.23, the FDDI standards specify the MAC sublayer and the concrete layer. The physical layer itself is divided into two sublayers. The standards likewise specify the station management (SMT) protocols. The PMD (physical medium dependent) sublayer specifies the fiber to be used equally well as the optical sources and detectors. The specifications of PMD are summarized in Figure 3.23. It should be noted that vendors make alternative PMD products available. For instance, twisted pairs can be used to connect stations separated by less than 100 m.

The PHY (concrete) sublayer specifies that the stations must use the 4B/5B encoding. With this encoding, the transmitter groups the bits past 4 and converts each iv-bit word into a 5-fleck discussion specified by the encoding table. The 16 words of v bits that the encoding uses were called and so that the resulting optical signal contains enough transitions to go on the receiver synchronized. Notation that with this encoding, the 100-Mbps data rates consequence in a raw bit stream of 125 Mbps on the fibers. If the transmitters had used Manchester encoding, the optical betoken would have transitions at 200 MHz, necessitating more expensive electronics.

The SMT must observe errors and isolate a fault on the ring, such as a failure of a station or link on the band. Figure three.24 illustrates how the dual ring is reconfigured equally a single ring after the error has been isolated. In addition, the SMT monitors the performance of the network. The MAC of FDDI specifies that the frames have a maximum length of 4,500 bytes. (The frame structure is illustrated in Figure 3.23.) The MAC uses a timed-token protocol. This protocol is similar to the token-passing mechanism of IEEE 802.5, except for the timing characteristic, as nosotros explicate next.

Figure 3.25 helps to explicate the MAC protocol when the stations transmit only asynchronous traffic. Assume that the stations are initially idle, that is, they have no packet to transmit. A token, which is a packet with a specific bit pattern, travels effectually the ring. Each station has 2 timers: TRT or token rotation time timer, which counts up, and THT or token property fourth dimension timer, which counts downwards. When a station has a package to transmit, it waits until information technology gets the token. When the station gets the token, it does the following:

i.: Grabs the token.
two.: Sets THT = TTRT − TRT. (TTRT or target token rotation time is fix by the network manager.)
iii.: Resets TRT = 0.
4.: Transmits packets until THT = 0 or at that place is no packet left.
5.: Releases the token.

Figure iii.25 shows for a particular station two successive token arrival and release times. Suppose that the fourth dimension, TRT, between successive arrivals is less than TTRT. The figure shows that in that case the fourth dimension betwixt successive token releases is also less than TTRT. Just this is the time between successive arrivals for the side by side station. By repeating the statement, nosotros conclude that every station volition wait for fourth dimension at nigh TTRT for a token inflow. If we assume that a station may complete transmitting its electric current packet even when THT = 0, so this argument must exist slightly modified to conclude that each station waits at most TTRT + TRANS, where TRANS is the longest packet transmission time.

The transmitting station must remove its own packet from the ring: the station waits until it receives the parcel that it transmitted, that is, until it reads its own physical accost as the source address of the bundle, and it then removes the package by transmitting "idle" symbols instead of repeating the packet.

Actually, the MAC protocol provides for two types of traffic: asynchronous and synchronous. As will exist explained, the stations get to transmit their synchronous traffic at least every two TTRT seconds. For example, if the stations concord on a value TTRT = xx ms, then the stations that transmit synchronous traffic, say vox, go to transmit at least every xl ms. If the voice is encoded into a 64-Kbps stream, then the stations need merely exist able to buffer 40 × 10⁻ⁱⁱⁱ × 64 × 10ⁱⁱⁱ = 2,560 $.25 of voice.

We now explicate how the protocol accommodates synchronous traffic. The stations first request permission to transmit synchronous traffic. The network eventually decides which stations tin can transmit synchronous traffic, and it allocates a fraction of TTRT to each of those stations. The fractions add upward to 1. The protocol works as follows. When a station that can transmit synchronous traffic gets the token, information technology does and then for upwardly to the fraction of TTRT that it was allocated. It transmits asynchronous traffic as before, using the previously described protocol. When the stations use this protocol, they get the token at to the lowest degree once every 2 TTRT seconds.

As was explained in Figure three.25, the MAC results in a bounded medium access time that is suitable for synchronous traffic. Annotation, nevertheless, that a station does not access the medium exactly at periodic times. Thus, the FDDI MAC does non implement an isochronous transmission facility. It can also exist shown that the FDDI MAC provides a fair allocation of the bandwidth to the dissimilar stations for asynchronous traffic. (A fair resource allotment is 1 in which every station has the aforementioned probability of transmission access. The fairness of the FDDI MAC is non quite obvious from our description of the protocol.) Moreover, the FDDI MAC protocol is very efficient considering the overhead that information technology imposes does not increase when the stations accept many packets to send. FDDI was designed to support multimedia connections where the stations exchange video, audio, text, and data. FDDI is existence used to interconnect LANs.

In summary, by using a timed-token mechanism instead of an untimed token-passing protocol or CSMA/CD, FDDI guarantees a divisional medium access time and is therefore suitable for synchronous transmission services in addition to asynchronous transmissions. Thus, the faster physical layer of FDDI increases the throughput, and the timed-token mechanism of its MAC enables the transport of constant chip rate traffic. Moreover, FDDI is designed to exist reliable against link or node failures. In 1999, it appears that FDDI is on the way out and is being replaced past switched Ethernet and, in some cases, by ATM as an interconnection technology for Ethernets.

Read full chapter

URL:

https://world wide web.sciencedirect.com/science/commodity/pii/B9780080508030500083

Networking Sensors

Feng Zhao , Leonidas J. Guibas , in Wireless Sensor Networks, 2004

3.2.ane The S-MAC Protocol

The principal goal of the Southward-MAC protocol is to reduce energy waste product caused past idle listening, collisions, overhearing, and control overhead. The protocol includes four major components: periodic heed and sleep, collision abstention, overhearing avoidance, and message passing.

Periodic listen and sleep is designed to reduce energy consumption during the long idle time when no sensing events happen, by turning off the radio periodically. To reduce latency and control overhead, Southward-MAC tries to coordinate and synchronize sleep schedules amid neighboring nodes by periodic (to compensate for clock drift) exchanges of the nodes' schedules, so that slumber times will be synchronized whenever possible.

Collision abstention in Southward-MAC is similar to the distributed coordinated role (DCF) for IEEE 802.11 ad hoc mode, using an RTS/CTS exchange. If a node loses in contention for the medium, it goes to sleep and wakes upward when the receiver is free and listening over again. The node knows how long to slumber, considering a elapsing field in each packet indicates how long the remaining manual volition be. Thus overhearing abstention is accomplished by putting nodes to sleep while their neighbors are talking to each other.

Messages are treated as logical data units passed between sensor nodes. In particular, a long message is fragmented into packets and sent in a burst with one RTS/CTS exchange to reserve this medium for the entire message. This saves repeated RTS/CTS overhead and reduces overall message-level latency. It does mean that a short bulletin may have to expect a long time while a long message finishes transmission, but every bit nosotros remarked, node-level fairness is non so important in sensor networks.

Because of such measures targeting improved energy efficiency, the energy consumption gap between 802.11-like protocols and South-MAC becomes significantly wider as the message interarrival period increases. Therefore, for an ad hoc sensor network with nodes remaining largely inactive for long times, South-MAC has obvious advantages over the 802.11 MAC in supporting typical sensor network applications today.

Read full chapter

URL:

https://www.sciencedirect.com/science/commodity/pii/B9781558609143500031

Mobile Wireless Network Security

Michel Barbeau , in Handbook on Securing Cyber-Physical Critical Infrastructure, 2012

5.2.3 Stream Cipher

At the level of WEP, a message is chosen a Medium admission command Protocol Data Unit (MPDU). Each MPDU is sent together with an Integrity Bank check Vector (ICV). It is an mistake detection code that has goose egg to do with security. CRC-32 is used for that purpose. However, if a MPDU is corrupted by communication errors, then very probable the errors volition be detected because of the inconsistency they create between the bulletin content and its ICV. When errors are detected, the MPDU is not accepted.

The stream cipher of WEP is illustrated in Effigy 5-1. WEP is analogous to a random number generator. In this instance, the random number generator is RC4. It generates a string of random $.25, as long as the length of the information to be encrypted and communicated, i.e., the length of the MPDU plus the length of the ICV. Both the source and destination of the MPDU use the same random number generator seed, determined jointly by a shared cloak-and-dagger primal and a public value, and do generate the aforementioned scrap cord. The encrypted message, by the source, results from the application of the xor performance on the $.25 of its MPDU, ICV, and bits of the random string. The destination recovers the text of the MPDU and error detection code by a 2d awarding of the xor operation on the encrypted MPDU, ICV, and the random cord. Because of the properties of the logic xor operator, the outset application is canceled, and the MPDU plaintext and ICV are returned.

The seed is constructed using a symmetric key (the source and destination own the same fundamental), and a serial number is assigned to each message called the initialization vector (Iv). It is similar if each message had its own new encryption seed. At that place are 2 key distribution models: default cardinal and central-mapping primal. A default central is a 40-bit or 104-bit symmetric key shared between an access point (AP) and several stations. Information technology is too called a network fundamental. A key-mapping primal is a 40-bit or 104-chip symmetric primal shared between an AP and one station. This is a pairwise key. The use of a 24-scrap IV and a 40-scrap key is called 64-chip security. The utilize of a 24-bit IV and a 104-fleck key is called 128-bit security. At a given station, all messages are encrypted using the same cardinal.

The IV is sent with the MPDU. Of form, the primal is not sent, merely an identifier of the primal that was used to encrypt the MPDU is sent. So in case, the source and destination share multiple keys, the destination tin can determine exactly which key the source used. A bang-up advantage of the stream zilch of WEP is the ease of implementation. It does non employ circuitous operations. At the time of its inception, it was possible to implement it in the wireless interface hardware.

The length of the IV field used by WEP turned out to exist as well curt. At today's current wireless information rate, very apace different MPDUs get encrypted using exactly the same Iv value, and when the same primal is used, they are encrypted with the same random scrap string.

Collection of several messages encrypted with the same IV and cardinal tin be done chop-chop. A 24-flake IV field ways that at that place are ii²⁴ (almost 17 million different IV values). For case, assuming a 100-Mbps network and a i-KB frame size (or 8192 bits/frame), the frame rate is equal to

$\frac{100 Mbps}{8192 bits/frame} = \frac{100 \times {ten}^{6} bps}{8192 bits/frame} = 12, 207 frames/s$

A full of 12,207 letters are transmitted per second over the network. If nosotros divide the number of possible Iv values by the frame charge per unit, it yields the maximum time it takes for IV values to get-go repeating

$\frac{17 \times 10^{6} 4 values}{12, 207 frames/s} = i, 393 s$

Information technology takes approximately forty min to use all possible IV values. 4 values are reused by individual stations and beyond stations (there is nothing unique to each station entering in the germination of the seed). Two messages encrypted with the aforementioned chip cord tin can be xor-ed together, and the result of the functioning is the xor of the plaintext of the two original messages. One may happen to know a few bytes of 1 of the ii messages. For example, if ane of the messages belongs to the Address Resolution Protocol (ARP), the message format is fixed and standard. Some of the field values are easy to judge. If the plaintext of one bulletin is xor-ed with the xor of the plaintext of the two original messages, and then the plaintext of the second message is obtained. This logic is at the basis of the WEP breaking procedure. Tews and Brook have shown that they can obtain the key within less than sixty due south [12].

Because of the flawed WEP, TKIP has been introduced as a replacement (come across Figure v-two). It is also a stream cipher based on RC4, but the IV field is longer. The RC4 seed is generated using something unique to each station, i.e., its MAC address. Ii frames encrypted with the same 4 are not encrypted using the same RC4 seed, and hence, they are not xor-ed with the same chip string. In contrast with WEP, the RC4 seed is calculated the same beyond all stations, particularly when a default key is used. 2 frames encrypted using the aforementioned IV and default key are encrypted using the same RC4 seed and are xor-ed with the same bit string. The WEP IV field is 24 bits, and the TKIP IV field is 48 bits (annotation that it occupies 56 bits, simply the seventh byte is not used). The seed is synthetic, by a two level mixing function, using a pairwise secret key, an IV value, and hardware address of the sender. Two stations with the same key and IV value very likely won't generate the same seed because they ain different hardware addresses and share different pairwise keys with the AP. With TKIP, 2 MPDUs with identical IVs but from dissimilar stations cannot exist xor-ed together to obtain the xor of the plaintext of the two letters.

Each MPDU is sent together with a 48-bit sequence number. It is used for replay protection. There is a 64-bit ICV that is based on the MICHAEL Bulletin Integrity Lawmaking (MIC) technique. MICHAEL computes a value using a hash office consisting of shift and add together operations. Given an input value, a hash function calculates an arbitrary value that doesn't seem related to the input. It is one mode if, given a value from the target domain, it is computationally hard to become back to the corresponding value in the input domain. Note that it is more than an error detection mechanism. A secret key is involved in the calculation (meet Department 5.two.five). It is truly a security mechanism that tin can notice bulletin tampering attacks. Devices supporting TKIP are eligible to the Wi-Fi Protected Admission (WPA) security certification.

A brute force assail can exist attempted to break TKIP. Frames are captured to get nonces entering in the calculation of the key during the four-style handshake for authentication and key establishment (see Section 5.2.5). Re-keying tin be forced using a de-hallmark assault. The animal force attack attempts to find the encryption key by trying all possibilities. A brute force search has been implemented in a software called coWPAtty [thirteen]. For each endeavor, in that location are 4096-hash computations involved. The procedure can be sped up with the apply of a rainbow table, i.due east., pre-computed hash values. Rainbow tables are, however, generated as a role of the Service Prepare Identifier (SSID) of the network beingness attacked. If no rainbow table is available for a SSID (probable to happen if the network doesn't operate with default values), so hash values need to be generated. Theoretically, the fourth dimension complexity of a brute force attack on TKIP is $O (2^{128})$ . It can be very tiresome and take years!

A animate being force attack can be greatly accelerated if combined with a lexicon attack. Entries in the dictionary are tried offset as potential keys. This works if the network is configured with a anticipated key. The pre-shared fundamental of TKIP is generated with a passphrase of up to 63 characters long, which tin be alphanumeric and punctuation characters. If all characters are used and selected randomly, and so a lexicon attack is very unlikely to succeed.

To the best of our knowledge, TKIP has not been croaky, merely weaknesses have been reported. Moen et al. have shown that information technology is theoretically possible to resolve the 128-bit temporal fundamental from TKIP RC4 encrypted frames [14]. The time complexity of their attack is lower than the brute force attack, simply still on the high side, i.e., $O (2^{one 05})$ , and may not be significantly more practical than the latter. Tews and Beck [12] have developed a Chopchop kind of assail to TKIP-encrypted ARP packets. The assail doesn't recover the temporal key, but decrypts the content of an ARP bundle. The attack requires repeatedly resending to the AP, the same parcel that needs to be decrypted. The content of the parcel is guessed byte by byte starting from the last. The reaction of the AP tells the attacker if he/she guessed correct. At this time, information technology is hard to claim that the level of difficulty of attacks on TKIP remains stiff. Progress has been made. Motivation is high among researchers to break TKIP because of potential recognition of peers. Information technology is likely that an eavesdropping assault on TKIP will eventually succeed. One needs to monitor the TKIP cracking progress.

Read full chapter

URL:

https://www.sciencedirect.com/science/commodity/pii/B9780124158153000054