Error detection/correction and fault detection/recovery – Pulse or data error handling – Error count or rate
Reexamination Certificate
2000-10-06
2004-01-06
Tu, Christine T. (Department: 2133)
Error detection/correction and fault detection/recovery
Pulse or data error handling
Error count or rate
C714S822000
Reexamination Certificate
active
06675328
ABSTRACT:
BACKGROUND OF THE INVENTION
Transmission of data with tolerable error rate in a communication network is desired. Generally, a communication system has the following components: a transmitter, a communication channel, a communication protocol, and a receiver. In most circumstances, each of these components is subject to independent design and testing specifications developed by multiple controlling organizations focusing narrowly on one element of the communication system. Also, manufacturers specializing in a particular component of the communication channel build the parts of a communication system. Thus, each of the components are typically designed, manufactured and tested independently of the other components in the system. Even if each of these components corresponds to its design and testing specifications, a communication channel assembled using the component might not perform adequately in terms of data throughput due to untested interactions among the components. In fact, most communications systems do not have any specifications and methods to test the entire system with all the actual components in place.
The International Standards Organization (ISO) has defined a layered model approach for communication networks, namely, the seven-layer model for open systems interconnection. The seven layers include the application, presentation, session, transport, network, data link, and physical layers. Each layer defines a different set of protocols necessary for communication, ranging from connectors and wires in the physical layer to identification of the reply address in an email in the application layer. Information always flows down through the layers from an application layer across a real physical connection between two or more physical layer objects.
In the highest layers (application oriented) and the lowest layers (physical or network hardware dependent) of the model, many standards are supported. However, the model is dependent on using essentially a single transport layer protocol to get all the different applications working with the different low-level networks.
Different protocols for the physical layer, for example 10BaseT or 100BaseT, can be used for data transmission. 10BaseT is a protocol that operates at 10 megabits per second (MBPS) and uses twisted-pair cabling. 100BaseT is a protocol for sending data over copper cables at a rate of 100 MBPS. Various standards govern a 100BaseT system that are different than those that apply to a 10BaseT system. The 100BaseT protocol is defined by IEEE Std 803.3u(1995), “Carrier Sense Multiple Access with Collision Detection (CSMA/CD) Access Methods-Type 100Base-T.” The main components of a 100BaseT system include data source coding, framing and sharing of the cable as specified in IEEE Std 802.3u. The coding and electrical specifications of the transmitted signal and the transmitter are specified in ANSI X3.263(1995), “Fiber Distributed Data Interface (FDDI)—Twisted Pair Physical Layer Medium Dependent (TP-PMD).” The specifications of the cable (for example, maximum length, impedance levels, attenuation and cross-talk) are specified by the standards TIA/EIA-568-A and TIA/EIA TSB 67. These specifications also describe connectors and patch cables. The receiver electrical and decoding specifications are described in ANSI X3.263(1995), “Fiber Distributed Data Interface (FDDI)—Twisted Pair Physical Layer Medium Dependent (TP-PMD).”
Like other communication systems, a 100BaseT computer network system includes a transmitter, a receiver and a communication channel system that includes cables and connectors to interconnect the transmitter and the receiver. Even if the system components of a 100BaseT system meet their individual acceptance criteria, cable plant assembled using these components can produce unacceptable levels of data errors measured as the ratio of packets (frames) lost due to Cyclic Redundancy Check (CRC) errors or missed detection to the total packets transmitted.
Existing testing methods may test the cable and cable components. The cable, connectors and patch cables are typically tested individually by the manufacturer according to the specifications in the TIA/EIA specifications. Further, the cable plant can be tested during installation, when Cable installers verify that the installed cable and components meet the TIA/EIA specifications for an installed cable channel. Most of the testing is done using handheld meters with measuring devices attached to both ends of the cable, and is performed before the transmitter and receiver are attached to the channel.
The Network Interface Card (NIC) manufacturers test against the FDDI-TP-PMD specifications. In addition, in certain communication networks, bit error rate testers send a preprogrammed signal through the cable and test for bit errors. However, the bit error rate testers test only the cable and do not test the actual transmitter, cable and receiver combination.
However, the end user of a communication system is more interested in the overall throughput of the system than in the performance specification of the individual components. Performance information becomes more useful if a certain throughput can be guaranteed by a communication system vendor; if the guaranteed throughput can be verified after the installation; or if the system throughput can be continuously monitored over the life of the communication system.
Using a specified communication throughput as a requirement also helps the communication system component manufacturers design and improve their products to achieve a better information throughput rate. For example, a cable manufacturer can design and optimize its cable for a particular communication protocol, transmitters and receivers. Likewise, a communication protocol designer can design a protocol that works best for existing components.
Bit error rate (BER) and packet or frame error rate are both measurements of system throughput. Frame error rate includes the ratio of packets of user information received without errors to the packets of user information sent. Bit error rate includes the ratio of quantum units (example: bits) of user or system information received without errors to the quantum units of user or system information sent.
Systems do exist today that are able to measure transmitter parameters, channel parameters, and noise levels. However, no system exists that combines the readily measurable parameters to produce a predicted or actual throughput measurement. Instead, existing systems are only capable of performing empirical throughput measurements. In such systems, a signal received by the receiver contains error-detecting information. Start and end codes and cyclic redundancy codes are incorporated within the data streams. Some Network Interface Cards (NICs) are capable of detecting errors using these self-consistency checks. By counting the number of errors received against a number of bits received, a bit error rate or packet error rate may be empirically calculated.
However, Empirical measurements of BER using a physical receiver are impractical for many reasons. First, typical error rates occur on the order of one error in 10
12
transmitted bits. At maximum transmission rate on a 100BaseT network, this is approximately one error in 2.8 hours. To empirically obtain a meaningful and statistically accurate error rate from a real system, one would need to measure for a duration of ten to one hundred times the error interval, so testing would require anywhere from one to ten days of measurement. Second, the physical hardware being tested is affected by a variety of external factors that cause inconsistent empirical measurements over repeated trials. This makes the calculation of an absolute BER of a communication channel (not including the receiver) impossible. Third, many physical receivers (i.e. NICs) provide only frame/packet error information. Any calculation of BER from the frame/packet error information is therefore approximate, since a packet error can correspond to more than one bit error. Finally, in many communicati
Judell Neil
Krishnamachari Srivathsan
Chase Shelly A
Rader & Fishman & Grauer, PLLC
Tu Christine T.
Vigilant Networks LLC
LandOfFree
System and method to determine data throughput in a... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with System and method to determine data throughput in a..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and System and method to determine data throughput in a... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3185595