Parallel File Systems

In cluster environments and MPPs, some file systems have been optimized to take advantage of the processor and memory resources represented by the nodes forming the cluster or MPP.

IBM’s General Parallel File System (GPFS) is an example of a parallel file system; it can be used on AIX clusters (HACMP), MPPs (IBM SP) or Linux clusters. Our description of GPFS is based on [HEG01]. GPFS’s major characteristics are:

» A clusterized file management system allowing transparent cluster file access (that is, program running on any node transparently accesses files, even if they are stored on another node)

» Scalability: GPFS has the ability to make use of processor, memory (used as a disk cache), and I/O resources of the nodes

» Failure-tolerant: GPFS provides journaling for metadata changes and data replication

In the SP implementation, GPFS is built on a software layer called Virtual Shared Disk (VSD). VSD allows disk blocks to be routed over a network, either an IP network or the interconnect network of an MPP. To this extent, VSD can be looked upon as a form of SAN (Storage Area Network), which we will discuss later.

GPFS is installed on the system nodes; it is possible to configure some nodes as specialist storage nodes. Data is shared by the applications running on the nodes provided with GPFS instances. Data is cached on the client nodes.

GPFS distributes data over the available disk, providing an effect similar to data-striping, which we will discuss later in the section on RAID systems.

Apart from AIX, the major software components of this architecture are the components of PSSP (Parallel System Support Programs), which are specific to the SP environment. The major elements are:

» VSD: Virtual Shared Disk, which provides the ability to access logical volumes as if they were local to the accessing node.

» GS: Group Services, which provides notification on the event of failure of a node or process, along with recovery of programs executing on the failing nodes on surviving ones. These services also initialize information necessary to VSD’s operation.

» RVSD: Recoverable Virtual Shared Disk, which makes it possible to prevent access by a node to certain disks during recovery phases of the node.

AIX also includes a component called VFS (Virtual File System), which allows applications’ file access requests to be directed to the appropriate file system (e.g., JFS (AIX’s journaled file system) or GPFS) transparently, depending on the type of the file.

Source of Information : Elsevier Server Architectures

IEEE 802.11n

In response to growing market demand for higher-performance WLANs, the IEEE formed the task group 802.11n. The scope of this task group is to defi ne modifi cations to the PHY and MAC layer to deliver a minimum of 100 Mbps throughput at the MAC service AP (SAP).

802 .11n employs an evolutionary philosophy reusing existing technologies where practical, while introducing new technologies where they provide effective performance improvements to meet the needs of evolving applications. Reuse of legacy technologies such as OFDM, FEC coding, interleaving, and quadrature amplitude modulation mapping have been maintained to keep costs down and ease backward compatibility.

There are three key areas that need to be considered when addressing increases in WLAN performance. First, improvements in radio technology are needed to increase the physical transfer rate. Second, new mechanisms implementing the effective management of enhanced PHY performance modes must be developed. Third, improvements in data transfer efficiency are needed to reduce the improvements achieved with an increase in physical transfer rate.

The emerging 802.11n specification differs from its predecessors in that it provides for a variety of optional modes and configurations that dictate different maximum raw data rates. This enables the standard to provide baseline performance parameters for all 802.11n devices, while allowing manufacturers to enhance or tune capabilities to accommodate different applications and price points. WLAN hardware does not need to support every option to be compliant with the standard.

The first requirement is to support an OFDM implementation that improves upon the one
employed in 802.11a/g standards, using a higher maximum code rate and slightly wider bandwidth. This change improves the highest attainable raw data rate to 65 Mbps from
54 Mbps in the existing standards.

Multi -input, multi-output (MIMO) technology is used in 802.11n to evolve the existing OFDM physical interface presently implemented with legacy 802.11a/g. MIMO harnesses multipath with a technique known as space-division multiplexing (SDM). The transmitting WLAN device splits a data stream into multiple parts, called spatial streams, and transmits each spatial stream through separate antennas to corresponding antennas on the receiving end. The current 802.11n provides for up to four spatial streams, even though compliant hardware is not required to support that many.

Doubling the number of spatial streams from one to two effectively doubles the raw data rate. There are trade-offs, however, such as increased power consumption and, to a lesser extent, cost. The 802.11n specification includes an MIMO power-save mode, which mitigates power consumption by using multiple paths only when communication would benefit from the additional performance. The MIMO power-save mode is a required feature in the 802.11n specification.

There are two features in the specification that focus on improving MIMO performance: (1) beam-forming and (2) diversity. Beam-forming is a technique that focuses radio signals directly on the target antenna, thereby improving range and performance by limiting interference. Diversity exploits multiple antennas by combining the outputs of or selecting the best subset of a larger number of antennas than required to receive a number of spatial streams. The 802.11n specification supports up to four antennas.

Another optional mode in the 802.11n effectively doubles data rates by doubling the width of a WLAN communications channel from 20 to 40 MHz. The primary trade-off is fewer channels available for other devices. In the case of the 2.4-GHz band, there is enough room for three nonoverlapping 20-MHz channels. A 40-MHz channel does not leave much room for other devices to join the network or transmit in the same air space. This means intelligent, dynamic management is critical to ensuring that the 40-MHz channel option improves overall WLAN performance by balancing the high-bandwidth demands of some clients with the needs of other clients to remain connected to the network.

One of the most important features in the 802.11n specifi cation to improve mixed-mode performance is aggregation. Rather than sending a single data frame, the transmitting client bundles several frames together. Thus, aggregation improves effi ciency by restoring the percentage of time that data is being transmitted over the network.

The 802.11n specification was developed with previous standards in mind to ensure compatibility with more than 200 million Wi-Fi (802.11b) devices currently in use. An 802.11n AP will communicate with 802.11a devices on the 5-GHz band as well as 802.11b and 802.11g hardware on 2.4-GHz frequencies. In addition to basic interoperability between devices, 802.11n provides for greater network efficiency in mixed mode over what 802.11g offers.

Because it promises far greater bandwidth, better range, and reliability, 802.11n is advantageous in a variety of network configurations. And as emerging networking applications take hold in the home, a growing number of consumers will view 802.11n not just as an enhancement to their existing network, but as a necessity. Some of the current and emerging applications that are driving the need for 802.11n are voice over IP (VoIP), streaming video and music, gaming, and network attached storage.

Source of Information : Elsevier Wireless Networking Complete

IEEE 802.11b — High-Rate DSSS

In September 1999, IEEE ratified the 802.11b high-rate amendment to the standard, which added two higher speeds (5.5 and 11 Mbps) to 802.11. The key contribution of the 802.11b addition to the WLAN standard was to standardize the PHY support to two new speeds, 5.5 and 11 Mbps. To accomplish this, DSSS was selected as the sole PHY technique for the standard, since frequency hopping (FH) cannot support the higher speeds without violating current FCC regulations. The implication is that the 802.11b system will interoperate with 1 and 2 Mbps 802.11 DSSS systems, but will not work with 1 and 2 Mbps FHSS systems.

The original version of the 802.11 specifi es in the DSSS standard an 11-bit chipping, called a Barker sequence, to encode all data sent over the air. Each 11-chip sequence represents a single data bit (1 or 0), and is converted to a waveform, called a symbol, that can be sent over the air. These symbols are transmitted at a one million symbols per second (Msps) rate using binary phase shift keying (BPSK). In the case of 2 Mbps, a more sophisticated implementation based on quadrature phase shift keying (QPSK) is used. This doubles the data rate available in BPSK, via improved efficiency in the use of the radio bandwidth.

To increase the data rate in 802.11b standard, advanced coding techniques are employed. Rather than the two 11-bit Barker sequences, 802.11b specifi es complementary code keying (CCK). CCK allows for multichannel operation in the 2.4 GHz band by using existing 1 and 2 Mbps DSSS channelization schemes. CCK consists of a set of 64 8-bit code words. As a set, these code words have unique mathematical properties that allow them to be correctly distinguished from one another by a receiver even in the presence of substantial noise and multipath interference. The 5.5 Mbps rate uses CCK to encode four bits per carrier, while the 11 Mbps rate encodes eight bits per carrier. Both speeds use QPSK modulation and a signal at 1.375 Msps. This is how the higher data rates are obtained.

To support very noisy environments as well as extended ranges, 802.11b WLANs use dynamic rate shifting, allowing data rates to be automatically adjusted to compensate for the changing nature of the radio channel. Ideally, users connect at a full 11 Mbps rate. However, when devices move beyond the optimal range for 11 Mbps operation, or if substantial interference is present, 802.11b devices will transmit at lower speeds, falling
back to 5.5, 2, and 1 Mbps. Likewise, if a device moves back within the range of a higherspeed transmission, the connection will automatically speed up again. Rate shifting is a PHY mechanism transparent to the user and upper layers of the protocol stack.

Source of Information : Elsevier Wireless Networking Complete

Security of IEEE 802.11 Systems

The IEEE 802.11 provides for MAC access control and encryption mechanisms. Earlier,
the wire line equivalent privacy (WEP) algorithm was used to encrypt messages. WEP uses a Rivest Cipher 4 (RC4) pseudo-random number generator with two key structures
of 40 and 128 bits. Because of the inherent weaknesses of the WEP, the IEEE 802.11i committee developed a new encryption algorithm and worked on the enhanced security and authentication mechanisms for 802.11 systems.

For access control, ESSID (also known as a WLAN service area ID) is programmed into each AP and is required in order for a wireless client to associate with an AP. In addition, there is provision for a table of MAC addresses called an access control list to be included in the AP, restricting access to stations whose MAC addresses are not on the list.

Beyond layer-2, 802.11 WLANs support the same security standards supported by other 802 LANs for access control (such as network operating system logins) and encryption (such as IPSec or application-level encryption). These higher-level technologies can be used to create end-to-end secure networks encompassing both wired LAN and WLAN components, with the wireless piece of the network gaining additional security from the IEEE 802.11i feature set.

Source of Information : Elsevier Wireless Networking Complete

WLAN Technologies

The technologies available for use in a WLAN include IR, UHF (narrowband), and SS implementation. Each implementation comes with its own set of advantages and limitations.


IR Technology
IR is an invisible band of radiation that exists at the lower end of the visible electromagnetic spectrum. This type of transmission is most effective when a clear line-of-sight exists between the transmitter and the receiver.

Two types of IR WLAN solutions are available: diffused-beam and direct-beam (or line-of-sight). Currently, direct-beam WLANs offer a faster data rate than the diffused-beam networks. Direct-beam is more directional since diffused-beam technology uses reflected rays to transmit/receive a data signal. It achieves lower data rates in the 1 – 2 Mbps range.

IR is a short-range technology. When used indoors, it can be limited by solid objects such
as doors, walls, merchandise, or racking. In addition, the lighting environment can affect
signal quality. For example, loss of communication may occur because of the large amount of sunlight or background light in an environment. Fluorescent lights also may contain large amounts of IR. This problem may be solved by using high signal power and an optimal bandwidth filter, which reduces the IR signals coming from an outside source. In an outdoor environment, snow, ice, and fog may affect the operation of an IR-based system.


UHF Narrowband Technology
UHF wireless data communication systems have been available since the early 1980s.
These systems normally transmit in the 430 – 470 MHz frequency range, with rare systems using segments of the 800 MHz range. The lower portion of this band — 430 – 450 MHz — is referred to as the unprotected (unlicensed), and 450 – 470 MHz is referred to as the protected (licensed) band. In the unprotected band, RF licenses are not granted for specific frequencies and anyone is allowed to use any frequencies, giving customers some assurance that they will have complete use of that frequency.

Because independent narrowband RF systems cannot coexist on the same frequency, government agencies allocate specific RFs to users through RF site licenses. A limited amount of unlicensed spectrum is also available in some countries. In order to have many
frequencies that can be allocated to users, the bandwidth given to a specific user is very small.

The term narrowband is used to describe this technology because the RF signal is sent in a very narrow bandwidth, typically 12.5 or 25 kHz. Power levels range from 1 to 2 W for
narrowband RF data systems. This narrow bandwidth combined with high power results in larger transmission distances than are available from 900 MHz or 2.4 GHz SS systems, which have lower power levels and wider bandwidths. Table 5.4 lists the advantages and disadvantages of UHF technology.

Many modern UHF systems are synthesized radio technology. This refers to the way channel frequencies are generated in the radio. The crystal-controlled products in legacy UHF products require factory installation of unique crystals for each possible channel frequency. Synthesized radio technology uses a single, standard crystal frequency and drives the required channel frequency by dividing the crystal frequency down to a small value, then multiplying it up to the desired channel frequency. The division and multiplication factors are unique for each desired channel frequency, and are programmed into digital memory in the radio at the time of manufacturing. Synthesized UHF-based solutions provide the ability to install equipment without the complexity of hardware crystals. Common equipment can be purchased and specific UHF frequency used for each device can be tuned based upon specific location requirements. Additionally, synthesized UHF radios do not exhibit the frequency drift problem experienced in crystal-controlled UHF radios.

Modern UHF systems allow APs to be individually configured for operation on one of the several preprogrammed frequencies. Terminals are programmed with a list of all frequencies used in the installed APs, allowing them to change frequencies when roaming. To increase throughput, APs may be installed with overlapping coverage but use different frequencies.


Spread Spectrum Technology
Most WLANs use SS technology, a wideband RF technique that uses the entire allotted
spectrum in a shared fashion as opposed to dividing it into discrete private pieces (as with
narrowband). The SS system spreads the transmission power over the entire usable spectrum. This is obviously a less efficient use of the bandwidth than the narrowband approach. However, SS is designed to trade off bandwidth efficiency for reliability, integrity, and security. The bandwidth trade-off produces a signal that is easier to detect, provided that the receiver knows the parameters of the SS signal being broadcast. If the receiver is not tuned to the right frequency, a SS signal looks like background noise.

By operating across a broad range of radio frequencies, a SS device could communicate clearly despite interference from other devices using the same spectrum in the same physical location. In addition to its relative immunity to interference, SS makes eavesdropping and jamming inherently difficult.

In commercial applications, SS techniques currently offer data rates up to 2 Mbps. Because the FCC does not require site licensing for the bands used by SS systems, this technology has become the standard for high-speed RF data transmission. Two modulation schemes are commonly used to encode SS signals: direct sequence SS (DSSS) and frequency-hopping SS (FHSS).

FHSS uses a narrowband carrier that changes frequency in a pattern known to both transmitter and receiver. Properly synchronized, the net effect is to maintain a single logical channel. To an unintended receiver, FHSS appears to be a short-duration impulse noise.

DSSS generates a redundant bit pattern for each bit to be transmitted. This bit pattern is called a spreading code . The longer the code, the greater the probability that the original data can be recovered (and, of course, the more bandwidth will be required). To an unintended receiver DSSS appears as low-power wideband noise and is rejected by most narrowband receivers.

Source of Information : Elsevier Wireless Networking Complete

WLAN Equipment

There are three main links that form the basis of the wireless network. These are:

● LAN adapter : Wireless adapters are made in the same basic form as their wired counterparts: PCMCIA, Card bus, PCI, and USB. They also serve the same function, enabling end-users to access the network. In a wired LAN, adapters provide an interface between the network operating system and the wire. In a WLAN, they provide the interface between the network operating system and an antenna to create a transparent connection to the network.

● AP : The AP is the wireless equivalent of a LAN hub. It receives, buffers, and transmits data between the WLAN and the wired network, supporting a group of wireless user devices. An AP is typically connected with the backbone network through a standard Ethernet cable, and communicates with wireless devices by means of an antenna. The AP or antenna connected to it is generally mounted on a high wall or on the ceiling. Like cells in a cellular network, multiple APs can support handoff from one AP to another as the user moves from area to area. APs have a range from 20 to 500 m. A single AP can support between 15 to 250 users, depending on technology, configuration, and use. It is relatively easy to scale a WLAN by adding more APs to reduce network congestion and enlarge the coverage area. Large networks requiring multiple APs deploy them to create overlapping cells for constant connectivity to the network. A wireless AP can monitor movement of a client across its domain and permit or deny specific traffic or clients from communicating through it.

● Outdoor LAN bridges : Outdoor LAN bridges are used to connect LANs in different
buildings. When the cost of buying a fiber optic cable between buildings is considered, particularly if there are barriers such as highways or bodies of water in the way, a WLAN can be an economical alternative. An outdoor bridge can provide a less expensive alternative to recurring leased-line charges. WLAN bridge products support fairly high data rates and ranges of several miles with the use of line-of-sight directional antennas. Some APs can also be used as a bridge between buildings of relatively close proximity.

Source of Information : Elsevier Wireless Networking Complete

imode Versus WAP

imode is available only in Japan, whereas Europe and other big markets for 3G mobile service providers are completely WAP-based. In the United States, most service providers have chosen WAP.

The most basic difference between imode and WAP is the different graphic capabilities; imode only supports simple graphics, which is far more than what WAP allows. The imode packet-switched data network is more suited for transferring data than the WAP CS network.

Another major difference is the “ always-on ” capabilities of imode. Since users are not charged for the time they spend on-line, it is more convenient and also less expensive. Since there is no need to dial up before using the various IP-based services, e-mail becomes as instant as SMS. imode uses cHTML, a subset of HTML, while WAP uses WML, a subset of XML. cHTML, while certainly easier to develop from a web-designer standpoint, has its limitations. The downside of WML, on the other hand, is similarly obvious — currently a WAP gateway is required to translate between HTML and WML for almost every data transfer. On the other hand, since WML is derived from XML, it is much more extensible. XML allows for more dynamic content and various different applications. In the future, a WML-based service will be of more advantage than an HTML-based one. So while WAP may currently require more complicated technology, in the long run, it may enable the user to do more with his or her device.

Source of Information : Elsevier Wireless Networking Complete

imode

imode is a proprietary mobile ISP and portal service from NTT DoCoMo, Japan, with about 50 million subscribers. For imode, DoCoMo adopted the Internet model and protocol. imode uses compact HTML (cHTML) as a page description language. The structure of cHTML means that the user can view traditional HTML and imode sites can be inspected with ordinary Internet web browsers. This is in contrast to WAP, where HTML pages must be translated to WML. imode provides Internet service using personal digital cellular-packet (PDC-P) and a subset of HTML 3.0 for content description.

imode is a packet-switched service (always connected, as long as the user’s handset is reached by imode signal) which includes images, animated images, and colors. In imode, users are charged per packet of downloaded information. imode allows application/content providers to distribute software to cellular phones and also permits users to download appilets (e.g., games). imode uses packet-switched technology for the wireless part of the communication. The wired part of the communication is carried over TCP/IP.

Packet -switched services send and receive information by dividing messages into small blocks called packets and adding headers containing address and control information to each packet. This allows multiple communications to be carried on a communication channel, giving efficient channel usage with low cost. Dopa, DoCoMo’s dedicated data communications service, offers connections to location area network (LAN) and ISPs.

The mobile packet communications system has a network configuration in which a packet communications function is added and integrated into PDC, the digital system for portable and automobile telephones. DoCoMo has developed a data transmission protocol specific to imode. This protocol is used with the PDC-P system. The PDC-P network includes a mobile message packet gateway (M-PGW) to handle conversions between the two protocol formats. Connection between the imode server and the Internet uses TCP/IP. The imode server is regular web server which can reside at NTT DoCoMo or at the enterprise. DoCoMo has been acting as a 1717 portal and normally maintains the imode server.

imode relies on Internet security as provided by SSL/TLS and does not have the ability to
handle server-side authenticated SSL sessions. imode phones are preconfigured with root
collision avoidance keys from public key infrastructure. This will allow for establishment of a server-side authenticated SSL session between the imode device and imode server hosted by the enterprise. imode does not have the capability of handling client-side certifi cates which means that nonrepudiation is not possible with the current implementation of imode.

Applications of WAP

The first and foremost application of WAP is accessing the Internet from mobile devices.
This is already in use in many mobile phones. This application is gaining popularity daily,
and many web sites already have a WAP version of their site. An application, which is out, is sending sale offers to mobile customers through WAP. The user’s phone will be able to receive any sale prices and offers from the web site of a store.

Games can be played from mobile devices over wireless devices. This application has been implemented in certain countries and is under development in many others. This is an application which has been predicted to gain high popularity. Application to access time sheets and filing expenses claims via mobile handsets are currently being developed. These applications, when implemented, will be a breakthrough in the business world.

Applications to locate WAP customers geographically have been developed. Applications to help users who are lost or stranded by guiding them using their locations are under
consideration. WAP also provides short messaging, e-mail, weather, and traffic alerts based on the geographic location of the customer. These applications are available in some countries but will soon be provided in all countries. One of the biggest applications of WAP under consideration is banking from mobile devices.

These applications will be very popular if they are implemented in a secure manner. The mobile industry appears to be moving forward, putting aside the issues of network and air interface standards, and instead concentrating on laying the foundations for service development, regarded by many as the key driver to multimedia on the move and third generation mobile systems. From that point of view, in the near future WAP and Bluetooth will play fundamental roles.

Source of Information : Elsevier Wireless Networking Complete

Cloud storage is for blocks too, not just files

One of the misconceptions about cloud storage is that it is only useful for storing files. This assumption comes from the popularity of file...