mail.beetsoslo.com/the-only-living-witness-the.php Sharing between networks of the fixed-satellite service using non-geostationary satellites and other networks of the fixed-satellite service. Dedicated user digital satellite communications systems and their associated architectures. Interference criteria and calculation methods for the fixed-satellite service.
Sharing between feeder links of the mobile-satellite non-geostationary service in the band 5 MHz and the aeronautical radionavigation service in the band 5 MHz. Frequency sharing between systems in the fixed-satellite service and wireless digital networks around 5 GHz. Performance objectives of digital links in the fixed-satellite service for transmission of Internet or higher layer Protocol packets. Technical and operational characteristics of networks of the fixed-satellite service operating above GHz.
Technical and operational considerations relating to the advance publication, coordination and notification of fixed-satellite networks. Development of methodologies for the assessment of satellite unwanted emission levels before launch. Fixed-satellite service systems using very wideband spreading signals. Interference between satellite news gathering SNG carriers by unintentional access. Frequency sharing between the FSS and the space research service in the Support of the modernization of civil aviation telecommunication systems and the extension of telecommunication systems to remote and developing regions with current and planned satellite networks.
Performance objectives of digital links in the fixed-satellite and mobile-satellite services forming elements of the Next Generation Network. Availability of digital paths in mobile-satellite services. Performance objectives for digital fixed-satellite and mobile-satellite services with variable bit-rate paths. Use of operational facilities to meet power flux-density limitation under Article 21 of the Radio Regulations.
Receiving earth station antennas for the broadcasting satellite service. Digital techniques in the broadcasting-satellite service sound and television. Frequency sharing issues related to the introduction of the broadcasting-satellite service sound in the frequency range GHz. Sharing studies between high-definition television in the broadcasting-satellite service and other services. Spectrum management issues related to the introduction of the broadcasting-satellite service sound in the frequency range GHz.
Digital broadcasting of multiple services and programmes in the broadcasting-satellite service. Contributions of the mobile and amateur services and associated satellite services to the improvement of disaster communications. Technical and operational characteristics for packet network transmission in mobile-satellite services. Characteristics and operational requirements of radionavigation-satellite service space-to-Earth, space-to-space, Earth-to-space systems. Interactive satellite broadcasting systems television, sound and data. Broadcasting-satellite means for public warning, disaster mitigation and relief.
System architecture and performance aspects on integrated MSS systems. Interference protection ratios and minimum field strengths required in the land mobile services. Characteristics of equipment for the land mobile service between 30 and 6 MHz. Techniques and frequency usage in the amateur service and amateur-satellite service. Interference to the aeronautical mobile and aeronautical radionavigation services.
Consideration of the needs of developing countries in the development and implementation of IMT. Quality of service requirements in the land mobile service. Reference radiation patterns of point-to-point fixed wireless system antennas for use in sharing studies. Sharing criteria between the fixed and land mobile services in the frequency bands above about 0. Use of the mobile, amateur and the amateur-satellite services in support of disaster radiocommunications. Nomadic wireless access systems including radio local area networks.
Further development of the terrestrial component of IMT. Protection criteria for aeronautical and maritime systems. Reference radiation patterns of omnidirectional and sectoral antennas for the fixed and mobile services for use in sharing studies.
Technical characteristics and channelling requirements for adaptive HF systems. Radio-frequency arrangements for fixed wireless systems. Technical and operational characteristics for systems in the fixed service used for disaster mitigation and relief. Frequency sharing and compatibility between systems in the fixed service and systems in other services. Operation of short-range radiocommunication public access system supporting hearing aid systems.
Performance and availability objectives and requirements for fixed wireless systems, including packet-based systems. Technical and operational characteristics of the land mobile service in the frequency range GHz. Technical and operational characteristics of stations in the fixed service in the frequency range GHz. Technical and operational principles for HF sky-wave communication stations to improve the man-made noise HF environment. Operational and radio regulatory aspects for planes operating in the upper level of the atmosphere.
Coexistence analysis between foreign object debris detection systems operating in the frequency range 92 to GHz and earth exploration satellite service sensors in-band and in adjacent bands. Serial data transport mechanism for packetized data within a television production studio based on, and compatible with, Recommendations ITU-R BT. Universal transmitters and retransmitters for both analogue and digital terrestrial TV broadcasting. Polarization of emissions in the terrestrial broadcasting service.
Generic bit-rate reduction coding of digital video signals for production, for contribution, for primary and secondary distribution, for emission and for related applications. Bit-rate reduction coding of audio signals for broadcasting applications.
Protection requirements of broadcasting systems against interference from radiation caused by wired telecommunication systems, from emissions of industrial, scientific and medical equipment, and from emissions of short-range devices. File formats and transport for the exchange of audio, video, data and metadata materials in the professional television and large screen digital imagery LSDI environments. Objective picture quality parameters and associated measurement and monitoring methods for digital television images.
Characteristics of terrestrial digital sound broadcasting systems for reception by vehicular, portable and fixed receivers. Subjective assessment of small, medium and large impairments in sound quality. Conditions for a satisfactory television service in the presence of reflected signals. Methodologies for subjective assessment of audio and video quality.
This invention uses specific 3-D modeling and, therefore, allows highly accurate asset management and facilities tracking of actual installed equipment while simultaneously providing for network performance prediction, measurement, and design capabilities that exploit the exact physical dimensioning of the network. In addition, the invention simultaneously stores an inventory of important network-specific and equipment-specific characterizations of all objects used in the network, such as vendor, model number, network hardware type, operating system version, firmware and software type and version.
The hierarchical, tree based model of the network is termed the Layout View. The physically accurate, site-specific model of the network is termed the Site View, whereby the attributes of each device can be displayed, stored or printed by selecting a particular item or node within the 3-D environmental model. Further, network hardware and software components can be interactively replaced, removed, reconfigured or moved to a new location in real-time using either the Layout View or the Site View.
An example of some of the information contained in the Layout View, hierarchical layout of a data communications network is shown in FIG. In the figure, a tree structure is used to display all hardware in the network. Each node in the tree contains information which is used to track the true physical location, logical layout and electrical, optical and electromagnetic connections for the data communications network hardware as well as any version numbers and settings of software or firmware running on that network equipment and the known performance parameters of that equipment, including the device throughput, bandwidth, quality of service, bit error rate, packet error rate, frame error rate, dropped packet rate, packet latency, round trip time, propagation delay, transmission delay, processing delay, queuing delay, network capacity, packet jitter, bandwidth delay product and handoff delay time.
The Site View of the invention has a physically accurate, three-dimensional modeling capability to display all network devices in a site-specific model of the environment that the network is located in. That is, the preferred embodiment of the invention allows each modeled hardware and software device to be placed in a three-dimensionally accurate manner and to track attributes of that device relevant to data communications networks.
These specifications may include important device or network subsystem operating parameters, such as throughput, bandwidth, quality of service, bit error rate, packet error rate, frame error rate, dropped packet rate, packet latency, round trip time, propagation delay, transmission delay, processing delay, queuing delay, network capacity, packet jitter, bandwidth delay product and handoff delay time. As described below, the Site View supercedes prior art described in previous co-pending patent applications by Wireless Valley Communications, Inc by hereby considering the difficulties and solving data network prediction, design and optimization problems for more complicated data communication networks.
Specifically, this new invention considers physical, site-specific modeling techniques and performance prediction methods and design methods for data network systems, both wired and wireless, which have performance characteristics that are based on much more complicated physical factors than just radio signal strength, interference, or multipath alone. In particular, for data communication networks, many additional factors, which relate to particular network equipment or modem designs, such as packet size, equalizer deployment, modulation methodology, source and error coding methods, packet protocols, as well as the number of co-channel network users, the type of persistency used for packet retransmission, or the multipath propagation effects in a wireless system, provide additional factors that must be considered in the design of a communication network that is designed for data traffic as opposed to simply voice traffic.
One difficulty that today's network designer or network system administrator faces is that most networking equipment uses proprietary, non-public methods for implementing various network devices, and these methods vary by specific vendor. Thus, it is difficult to form reliable prediction models by just using basic physical propagation models in a wireless network, for example.
As data transmission technologies such as Bluetooth, DSL, Voice over IP, and future packet-based cellular radio network architectures proliferate, the ability to predict and measure specific network performance parameters will become increasingly important, and the ability to properly incorporate measurements into 3-D prediction models for performance parameters will be important for proper network deployment. This invention considers attributes relevant to packet-switched data communication networks, which require more extensive and non-obvious modeling when compared to traditional cell phone or telephone voice communication systems that are circuit switched and use a dedicated single user or bounded number of users per assigned operating channel.
Data communication networks have performance criteria that are specific to packet-based systems and that are not useful to all types of communication networks contemplated previously. For this reason, the preferred embodiment of the invention can additionally predict the throughput, bandwidth, quality of service, bit error rate, packet error rate, frame error rate, dropped packet rate, packet latency, round trip time, propagation delay, transmission delay, processing delay, queuing delay, network capacity, packet jitter, bandwidth delay product and handoff delay time, based on the specific physical and spatial location of each network component, as well as the physical, electrical, and logical attributes of the specific components.
The performance prediction methods take into account all devices and network equipment, including the physical locations within the 3-D modeled environment, using the constructed Bill of Materials of the network within the 3-D modeled environment, and is capable of performance predictions for any desired location in the modeled network and environment, where a location may be within a room, at a particular location in a room, within a building, or in an outdoor region of varying granularity, depending on the requirements of the user.
Prediction of throughput, bandwidth, quality of service, bit error rate, packet error rate, frame error rate, dropped packet rate, packet latency, round trip time, propagation delay, transmission delay, processing delay, queuing delay, network capacity, packet jitter, bandwidth delay product and handoff delay time and other performance parameters may be carried out by predicting the performance for all wired network components separately from the performance of wireless components, and then combining the results to get the net network performance.
To predict the performance of a wired communication link, it is important to combine the known effects of each piece of wired equipment for the specific network settings, also known as operating or performance parameters, such as protocol type, data type, packet size, and traffic usage characteristics, firmware type, operating system type, typical network performance characteristics, and typical, average, peak, and minimum traffic load on the network.
For wireless network components, additional factors concerning propagation, signal strength, interference, and noise must be considered. The preferred embodiment of the invention allows data communication networks to be accurately characterized for performance prediction in a number of novel ways.
First, performance prediction may be based on field measurements from an actual network, where prediction models are formed from some fit to measured data an empirically-based model. These field measurements may be made manually, or autonomously, using data collectors, or agents, that continually measure and update the specific network performance metrics that are observed within the physical environment. These data collectors are able to measure, or are assigned, specific 3-D position locations within the physical environment, such position locations corresponding to known positions in the computer model which is used to model the physical environment of the network, and which are known or which are transmitted to a measurement server.
The data collectors may be individuals who manually or automatically record or collect observed network performance such as one or more of the aforementioned performance parameters, or the measurement agents may be software or hardware or firmware applications that run on top of network applications for the purpose of routinely measuring for one of more of the numerous network performance parameters listed previously.
The agents may be fixed, or may be portable, and may have position location devices, such as GPS or inertial navigation, or an internal map which is activated by a user, so that the position location of the measurement is sent to a server processor. The agents are presumed to have two-way communication with a server processor that may be collocated or remotely located.
Measurements from one or more data collectors are routinely or periodically collected and then transmitted, either by wireless or wired means, or by real-time or stored means, to a server processor which is either collocated, or remotely located, from one or more of the measurement agents. For example, the measurements may be recorded by autonomous agents and then transmitted over a fixed network to a processor that integrates all measurements and computes statistics for observation. The measurement sources have known positions in 3-D, or may not be known and used to form a gross estimate of observed network performance.
The collected measurements may be sent in real time, stored and forwarded, or sent as file transfers via many means, such as via email, over the world wide web, via wireless, wired or optical links, or in a storage device. Using the measurement information from the data collectors, the server is able to provide a predictive model by using knowledge of the physical. In the preferred embodiment of the invention, the server stores and processes the physical location of all measurement devices where available as well as all network components and their electrical, logical and technical configuration, while also considering cost and maintenance issues associated with each network component.
Using the preferred embodiment, a data communications network can be designed, deployed, tested, predicted, measured, optimized and maintained by collecting the measured data from one or more agents, and processing them at the server to determine a proper prediction engine that allows future network layout with a desired outcome prior to installation.
The server engine is able to display the measured results, in a site-specific manner from each measurement agent that has site-specific information so that predictions may be compared to measurements on a visual display of a computer or in a stored means such as an ASCII file comparing predicted versus measured performance parameters. It is important to note that each measurement agent may be a server, capable of fusing measurement data with the site-specific 3-D layout of the network components and the physical environment.
Therefore, each measurement agent may serve as a centralized processor, as well, so that many different physical locations of a particular network may be measured and predicted for performance. Servers may then be collocated or remotely located from the measurement agents, which collect, display, store and use the measurements to form predictive models. In the case of a remote server that receives measurement data from measurement agents, it is possible to remotely monitor, and then predict, the performance of a network that is physically very far from the particular server processor.
The measurement agents may be further controlled or configured by the server processor, so that the agents may be tuned or instructed to perform different types of measurements, such as different packet transmission rates, observation intervals, averaging intervals, protocol types, or other sensible changes which those skilled in the are would conceive for proper network optimization. A second method for predicting the performance of network parameters is through the use of analytical or simulation methods.
These analytical and simulation methods are well known, and relate the physical and electrical characteristics of the network channel to the physical and electrical characteristics of the various network components. Through simulation or analysis, it is possible to determine approximations or bounds on the typical values that one would expect in an actual network configuration of specific components. The present embodiment of the invention allows a user to enter the results of such calculations, so that they are applied as inputs to the prediction model.
These first-guess values may then be iterated by the invention, based on feedback from the site-specific measurements of the actual network. A measured set of data for a typical operating environment with multiple transmitters in a wireless or wired network, are recorded, stored and displayed by the invention, as taught in the previous description about the measurement agents and server processors.
Then, some form of best-fit algorithm minimum mean square, median filter, etc. This table look up method allows measured data to be translated into values that may then be used to drive predicted data for all subsequent predictions conducted within the same site-specific 3-D environment in which measurements were made.
Alternatively, best guess performance metric values, or best guesses for the functions or constants in the equations listed below, may be fed into the invention, either manually or automatically through a storage means or via a wireless or wired means from a remote or collocated location, for a specific 3-D modeled network environment, wherein the predicted performance at any space or location with the 3-D environment is based on the first, best guess, predictive models. The empirically-based predictive models and the initial best guess predictive models may be used in subsequent environments, different from the environment for which measurements or best guesses were made, and the invention allows a catalogue of models to be used easily by the user for subsequent network prediction or design.
Measurements of actual network performance may then be overlaid and displayed and stored simultaneously with the network prediction parameters, for rapid comparison. Furthermore, optimization routines compute the best values for minimum error for new predictive models that match the measured network performance within the environment.
Thus, the invention allows the user to relate empirically-derived predicted performance parameters or initially guessed network performance parameters within a 3-D site specific configuration of the actual installed or contemplated network, using specific information and physical locations about the network devices and by using the models for wired networks and wireless propagation, multipath, and noise.
Thus, performance prediction can be ascertained and compared to measured network performance for use in ongoing network deployment. Furthermore, by comparing measured network performance metrics to predicted metrics, the invention allows new field measurements to update the previous prediction models in a convenient method, which provides a catalogue of models that is stored and displayed to the user either locally or remotely.
Alternatively, using the hierarchy of servers, it is possible to use remotely located servers which compute, transmit, or receive such measurements and predictive models for the remote use, display, measurement and storage of model parameters and results. This is particularly convenient for network administrators who wish to monitor the performance and design of networks that are physically distant from the network of interest. Measurements of a particular device for desired performance criteria is accomplished either by using the measurement software module available in the preferred invention or by importing a log file from another software or hardware measurement tool.
The measurement module within the preferred invention allows the measurement of the performance of any specific portion of a communications network using two or more software programs which are installed and run on either sides of a device or devices. These software programs are called agents. By sending test transmissions between two agents across a specific network connection the preferred invention can measure any particular performance criterion.
The results of these measurements are stored for a particular portion of the network. The preferred embodiment of the invention can also import the logfiles of other measurement programs such as traceroute to measure specific links. This functionality allows site-specific measurements made by external programs to be stored site-specifically.
This is accomplished by a two-pass method described in patent Ser. Skidmore, filed Dec. To import a logfile a user simply clicks a point in the model of the environment for each data point to assign a location for each point in the logfile. In performing network performance measurements, especially for wireless data networks, it is important to know the difference in performance for transmission and reception.
This is why the preferred invention can measure the transmission and reception components of the average network statistics. To measure the transmission direction, the size of test packets is varied. By changing the size of the packet sent and the size of the packet returned, the transmission and reception statistics can be separated. This allows a network designer to identify problems in transmission that might otherwise be masked by apparently good reception.
Network performance measurements are not useful if the measurements do not mimic the actual data traffic that a network carries. For this reason, the preferred embodiment of the invention is able to mimic the traffic patterns, network protocols and packet characteristics of actual data. Thus, if web browsing performance is being measured, the invention sends small packets from an access terminal to a web server and returns large packets from that server that are typical of text, image and web script file formats. By measuring the performance of such packets, the invention accumulates accurate network statistics for expected web browsing performance.
The measurements of specific traffic types may also be applied to the use of broadcast or multicast packet performance scenarios. The preferred embodiment of the invention is able to measure performance of multiple transmitters or multiple receivers or both of the same packet information. The performance of this type of transmission are different than point to point measurement because shared resources are used more efficiently in broadcast and multicast scenarios. Thus, the ability of the invention to measure network performance statistics for the overall success of the broadcast or multicast transmission and for each individual transmitter and receiver is quite powerful.
This ability allows network designers to better choose which transmitters of multicasts might be redundant or which broadcast transmissions are insufficient to reach all the desired receivers. In some data communications network, the performance of specific pieces of equipment, such as Ethernet Bridges or even a single cable, is hard to measure because it is transparent to the network layer of a data communications network.
For this reason, the ability of the invention to determine the performance of a single device through extrapolation is quite useful. The preferred embodiment of the invention is able to use known performance data for specific pieces of network equipment and extrapolate the contribution of other devices in the network. Measuring and extrapolating enough individual hardware and software links can identify the performance of all network devices.
The accuracy and reliability of this procedure heavily depends on an accurate and site-specific model of the data communications network, which the invention possesses. Extending the extrapolation concept of performance evaluation to the software and hardware components of network equipment demonstrates a further capability of the preferred embodiment of the invention; The invention is able to distinguish in some cases between the performance limits due to software and those due to hardware.
For example, in a situation where the transmitter and receiver are the same computer, no hardware is actually involved in the transmission. By measuring network statistics in this situation, one can quantify the performance of just the computer software. By comparing the situation where the transmitter and send are the same to a situation where the transmitter and receiver are different computers the performance of just the computer hardware can be identified.
Since the performance of the software in either case will be quite similar, the performance of just the hardware in a connection between two computers can be extrapolated by assuming the software will perform similarly in either case. Extrapolating the performance of individual network components from measured performance metrics can be time consuming.
For this reason, the preferred embodiment of the invention is able to read in data results from a plethora of measurement tools, system utilities and network logfiles to a single internal format.
The invention is capable of reading in the output of command line utilities such as ping or ttcp, the logfiles generated by routers and switches such as tcpdump, or even the logfiles of other commercial measurement programs, and these measurement results are stored for use in the predictive engine. The combination of these imported files to a single internal format allows the invention to combine many different measurements and activity logs into a single set of network statistics. This process means the invention requires fewer active measurement campaigns and more diverse and accurate data for better and more accurate network performance modeling.
Accurate, reliable representations of a data communication network require a large number of measured data points. Hence, the preferred embodiment of the invention collects a large amount of data quickly and easily using various methods as described above. The invention does this by providing remote data collection agents, which can be installed on data access terminals or embedded in hardware, software, or firmware within an actual device in the network. The remote data collection agents respond to a server program the processing server that controls the measurements made by the remote agent.
That is, the remote agent can be directed to make a measurement to or from any other remote agent or processing server using any desired protocol, traffic type, network setting, or configuration. This process does not require any input from a human user at the remote agent's physical location. The agents simply records the data when asked with the correct settings and reports the results back to a server which stores data from all remote agents and other measurement tools.
The server can generate a variety of detailed reports and use the data to make predictions about expected network performance in future. Servers can also function as agents. In this manner, servers can be organized in a hierarchy or a distributed fashion. This allows servers to report measurements to one another and make measurements using other agents or servers.
A network designer at a server can then use all collected and reported data to identify problem areas such as fairness or poor distribution of broadcast data, or problem times, such as increased network activity at lunch time with a data communications network. In order to improve the value of measurement data collected, the preferred embodiment of the invention identifies the exact if possible or approximate location of a remote agent. As discussed earlier, remote agents in this case can either be controlled by a user at that physical location, or controlled remotely by a server.
In the preferred embodiment of the invention, the agent uses information about the network layout to identify an approximate location.
Determining the nearest piece of network equipment and associating the approximate location with the precisely known location of that network equipment accomplishes this. This approximate location can be further refined using dead reckoning, clicking on a location in a map, or using the global positioning system, laser range finders or some other positioning device known now or in the future.
The preferred embodiment of the invention is not only capable of accounting for the effects of different hardware, firmware, software and configuration settings, but it can also predict the effects of just the hardware and firmware, just the software, or of a single configuration setting. The ability of the invention to measure and thus adjust empirically-derived predictions for these effects allows the optimization of the data communications network. By predicting the effects of changing any detailed aspect of the data communications network, a user can immediately visualize the effect of a new component or a setting change.
This ability allows a user skilled in the art to design an optimal data communications network by continually making changes and observing the prediction changes.
Growing demands of wireless local area communication are prominent for many of Spectrum Management Methods for Wireless Local Area Networks. Design and Simulation of Spectrum Management Methods for Wireless Local Area Networks by Andreas Könsgen, , available at Book.
We now focus on the details for predicting values for network performance parameters based on knowledge of the 3-D site-specific environment as well as the specific components used in the network design. The throughput and bandwidth of a network are calculated by the invention as functions of any or all of the following operational parameters which impact performance: distance between transmitter and receiver, physical environment specification, packet sizes, error and source coding schemes, packet overhead, modulation techniques, environment, interference, signal strength, number of users, and for wireless networks, the antenna pattern and type, multipath delay, number of multipath components, angle of arrival of multipath components, radio frequency bandwidth, protocol, coding scheme, and 3-D location.
In order to predict the bandwidth and throughput of a network connection, the appropriate functions and constants may be calculated from the listed parameters and then predicted for each location and time desired. For a wired network, throughput T or bandwidth BW may be derived from a vendor's specification sheet of a product or device, or may be measured in a special laboratory setting. Alternatively, T or BW may be calculated through analysis or simulation, or may be measured in the field using a number of known devices.
These means may be used to determine the proper value for T or BW in a network prediction enging such as contemplated here. A formula for predicting the throughput and bandwidth for a wireless data communications channel is shown in equation 1. RSSI is the received signal strength intensity, which is the power level of the signal at the receiver, either in absolute values or in logarithmic values. A, B, C, C 1 , C 2 , C 3 , D, E, F, K i , are constants or may represent linear or nonlinear functions of one or more physical or electrical parameters, such as physical environment type, packet size, modulation, modem type, or other parameters that relate the physical, electrical, or logical environment of the network.
These constants or functions take on specific functional values depending upon if T or BW is being solved for. The values of G i and P i represent gains and power levels, respectively, for each of M different signal components, which may represent individual multipath components or gross signal components from one or more radiating sources, and K i represents a finite number of constants or functions for each value of i.
Note that Gi, Pi, and the individual Ki may be in logarithmic e. It is important to note that multipath delay, and its effect on network performance prediction and design, may be considered in many ways, as contemplated by this invention and as shown in Equation 1. First, multipath may be considered individually, whereby each multipath component is considered to arrive from each transmitting device, and the methods for modeling multipath are well known and explained in the prior art, and in numerous research works by Rappaport, et. Alternatively, gross multipath effects may be modeled as having a worst-case delay e.
Alternatively, spatial considerations may be used by contemplating the antenna patterns of each transmitter or receiver, so that multipath which arrives only in the main beam of each wireless device is considered in the calculation of delay and in network performance in 1. Alternatively, only the strongest one or two or some finite number of transmitters may be considered for multipath propagation delays, whereby only a finite set of transmitters, such as those most closest to the receiver of interest, or those of a certain standard, frequency, or power setting, are considered to radiate multipath energy and produce RSSI values, and from that finite number of transmitters, only the strongest multipath, or the average, maximum, median, or largest few multipath components are considered in computation of delay.
Alternatively, if only a finite number of transmitters are considered, methods described above, such as consideration of the physical environment to determine a gross multipath delay from each transmitter, or the use of a particular antenna pattern to determine most important multipath components, may be used to drive the model of multipath and its impact on network performance. Similar approaches may be used to model the received signal strength, RSSI in equation 1. Note that the constants or functions of equation 1 may be assigned blindly for initial predictions, and then a specific network within the site-specific environment may be measured empirically so that a best-fit using a minimum mean square error approach or some other well known method may be used to assign values for the constants or functions in 1.
Propagation delay for network data is predicted for wired networks, where components are interconnected by wire either fiber or metal wire by dividing the distance traveled by the propagation speed of the electrical, electromagnetic or optical signals in the device, which are used to transmit the data. Such photons move at the speed of light in glass, which is less than the free space propagation speed.
By using the site-specific method of modeling the complete network within the present invention, it is possible for the user to simultaneously visualize the network as configured in the environment and see a display of delay and predicted or measured performance of delay within the cable within the 3-D environment. Additionally, using a tool tip mouse cursor or some other pointing means, or using a pull down menu, or by simply viewing the display device which the invention is implemented on, various network performance metrics, as well as stored data from the Bill of Materials and parameters of intere may be visualized or stored.
Predicting the propagation delay for a wireless portion of a data communications network is more difficult than wired networks due to the fact that multiple transmitter sources, such as access points in a Bluetooth network, IEEE Furthermore, as mentioned previously, multipath interference can create echoes that may or may not be equalized depending on the specific network equipment used at the wireless receiver or transmitter.
However, the same calculation model used for wired networks may be used, with the additional consideration of multipath delay terms, and propagation losses or gains, due to specific multipath components, as shown in Equation 1. This additional consideration of multipath delay is needed to account for the fact that wireless data does not always travel in a straight line, and that physical objects can diffract, reflect, absorb, and scatter radio energy. Predicting the multipath delay is performed using well-known raytracing techniques or based on angle of arrival, or signal strength values, or by making estimated based on the physical model of the 3-D environment.
Transmission delay is directly calculated from the bandwidth of a connection using the number of bits transmitted. To calculate transmission delay, the number of transmitted bits is divided by the bandwidth. This calculation is identical for wired and wireless channels but must be performed separately for each network device.
The formula is illustrated in equation 3. Processing delay must be calculated for each device separately within a network. Processing delay is the time required for a network device to process, store, and forward the data bits that are applied to a network device. Alternatively, processing delay may be the time required for a source to produce a meaningful data stream once it is instructed to do so. Processing delay is known to be zero for devices that do not perform any processing, such as passive network components like cables, antennas, or splitters.
Processing time may depend on the packet size, protocol type, operating system, vendor, firmware, hardware, and software versions or configurations, and the type of device and the current computing load on the device. To predict the processing delay of any device it is necessary use a model that accounts for all of these effects. These models may be measured in the field, measured in a test facility, obtained from vendors, or derived from analysis or simulation.
Queuing delay is only applicable to devices that transmit data from multiple users or multiple connections. The queuing delay of a device is the amount of time a particular packet must wait for other traffic to be transmitted. It is difficult to predict the queuing delay of a particular connection because it depends on the amount of traffic handled by a particular device. Alternatively, average, median, best or worst case, or some other linear or nonlinear weighting of queuing delay times as defined by the device specifications, or as measured, simulated, or computed by analysis, may be used to calculate a predicted queuing delay time.
Packet latency, round-trip times and handoff delay times are all based on propagation, transmission, processing, and queuing delay times. To accurately predict packet latency and round trip time, the propagation, transmission, processing and queuing delay times must be summed for all network devices in a particular network link and adjusted using the particular traffic type, packet size, and protocol type. For instance, packet latency is the time required for a packet to travel from transmitter to receiver.
To predict packet latency for a particular link the propagation, transmission, processing and queuing delay times must be calculated using the specific network connection, traffic type, and packet size for the one-way transmission of a packet. Round trip times are calculated similarly, except for the transmission and reception of a packet and the return of the acknowledging packet. Thus, to predict the round trip time, the invention takes into account the original packet size and the size of the acknowledging packet as well as the effects of the specific network connection, protocol and traffic type on the propagation, transmission, processing and queuing delays.
Handoff delay times are based on the propagation, transmission, processing and queuing delays involved in two separate wireless access points coordinating the change of control of a wireless device from one access point to another. These delays result because the two access points must transmit data back and forth to successfully perform a handoff. Thus, the prediction of handoff delay time is similar to the prediction of the packet latency time between the two access points. To predict the handoff delay time, the invention calculates the propagation, transmission, processing and queuing delays for the link between the two access points.
The invention then adjusts for the specific number of transmissions required and the size of the data, which must be sent to successfully perform a handoff. When predicting bit error rates, the invention considers wired and wireless error rates. Wireless networks operate in much more hostile electrical environments than their wired counterparts and their interconnections are significantly more difficult to model and, until this invention, practical networks have not successfully been modeled using specific, accurate physical and electrical models of multiple transmitters, multiple interferers, noise sources, and network components within a 3-D site-specific environment.
This invention uses 3-D site specific representations of the environment for specific network implementations that are able to consider both wired and wireless networks, and considers physical locations, electrical specifications and attributes of all radiating sources and their antenna systems in a real-world 3-D environmental model. Wireless networks are prone to data errors much more so than wired channels, due to the impact of multipath propagation, multiple transmitters, and noise, as described previously. The fact that radio propagation and noise is more random than for fixed wired networks must be considered for practical design, and is modeled in this invention.
For wired channels, bit error rates are simply a measure of the electrical, optical and electromagnetic parameters of a connection and are predicted using a statistical random variable, such as a Gaussian or Poisson random distribution, or other sensible distribution or algorithm known now or in the future, and this random variable is overlaid about the average, median, or typical performance of the network component or network subsystem.
The network may be wired, wireless, or a combination thereof. Many performance metrics of a device or a network subsystem, such as Frame Error Rate, Bit Error Rate, or Packet Error Rate, as well as other performance parameters such as throughput, bandwidth, quality of service, bit error rate, packet error rate, frame error rate, dropped packet rate, packet latency, round trip time, propagation delay, transmission delay, processing delay, queuing delay, network capacity, packet jitter, bandwidth delay product and handoff delay time may be either derived from a specification of the equipment, may be calculated analytically within the invention or inputted into the invention, or may be measured a priori in advance to using the invention.
That is, specific parameters of operation, known as operating parameters or equipment parameters, such as those listed previously, can be either measured or predicted through equipment specifications provided by vendors. Alternatively, they may be measured in-situ by a user or research facility, for proper modeling and input into the invention. Alternatively, they may be calculated based on some known analytical model that contemplates interconnection of devices so that a performance model and operating parameters maybe computed.
The statistical random variable to model network performance within the invention can be dependent on the electrical, optical and electromagnetic characteristics of each device such as voltage levels, power levels, impedance, and operating frequencies, or can be generated using a typical observed measured value for each network device.
For instance, copper wire can be modeled as having a bit error rate of 1 error in 10 6 or 10 7 bits transmitted. Once measured and characterized a single initial time, a single component or a string of components within a network may be modeled repeatedly by the invention, so that network performance models. Wireless performance parameters, however, are dependent on many more factors than wired bit error rates. For this reason, the invention predicts wireless bit error rates based on the environment, distance between transmitter and receiver, number and types of partitions obstructing the transmission, time, 3-D position, packet size, protocol type, modulation, radio frequency, radio frequency bandwidth, encoding method, error correction coding technique, multipath signal strengths and angle of arrival, and multipath delay.
As a result, the calculation of the predicted bit error rate is performed using constants or functions to convert from previously measured or known channel and network equipment performance metrics to an expected bit error rate. A formulation for predicting the bit error rate, frame error rate or packet error rate directly for a data communications channel is shown in equation 4, and is identical to equation RSSI is the received signal strength intensity, which is the power level of the signal at the receiver.
The each of M values of G i and P i represent gains and power levels, respectively, of different signal components, which may represent individual multipath components or gross signal components from one or more radiating sources, and may be in logarithmic or linear values of power. The variables G i and P i and each one of the M number of Ki values may be in logarithmic e.