Multiplex communications – Communication techniques for information carried in plural... – Transmission bandwidth conservation
Reexamination Certificate
1998-04-20
2002-08-20
Vu, Huy D. (Department: 2665)
Multiplex communications
Communication techniques for information carried in plural...
Transmission bandwidth conservation
C370S252000, C370S468000
Reexamination Certificate
active
06438141
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The invention relates generally to communication of information and, more particularly, to management of communication of information over media of finite bandwidth.
2. Background Art
Sun, Sun Microsystems, the Sun logo, Solaris, “Write Once, Run Anywhere”, Java, JavaOS, JavaStation, HotJava Views and all Java-based trademarks and logos are trademarks or registered trademarks of Sun Microsystems, Inc. in the United States and other countries.
The overall performance of a computing system having components that are physically separated from one another is often limited by the rate at which information may be moved between the components. The communication media for transferring the information between the components has a limit as to the amount of information that can be transferred over it in a given amount of time. This limit is referred to as the bandwidth of the communication media. This limit is, in some sense, analogous to the maximum amount of water that can be pumped through a particular size of pipe in a given amount of time. While water pumped through a pipe may be measured in gallons per minute, for example, the bandwidth of a communication medium is often expressed in bits per second, where each bit is a binary digit. A binary digit is a small unit of information that has either the value zero or the value one and from which other larger units of information may be constructed.
Furthermore, just as a cable television company may combine the signals from many television programs to be carried over one high-capacity cable, it is sometimes desirable to transmit information from multiple sources through a communication medium. In some configurations, multiple sources may transmit information to a single destination, for example, a single user terminal.
However, information passed between components of a computing system is often passed at an irregular rate. For example, much information might be transferred to send a document to a user terminal for display, but little information might be transferred while a user reads the document while it is displayed on the user terminal. Thus, the information may be passed in short bursts of large amounts of information followed by long periods of small amounts of information.
Moreover, each of the multiple sources transmitting information might transmit their respective information irregularly, with large variations between times of high transmission rates and times of low transmission rates. Thus, situations may occur where multiple sources attempt high transmission rates simultaneously. If the instantaneous demand for information transmission exceeds the bandwidth of the communication medium, not all of the information will be able to be passed through the communication medium at the rate it is being fed to the communication medium. Thus, a technique is needed to manage the flow of information through a communication medium and prevent such congestion of the medium from interfering with efficient transfer of information.
Additionally, it is not always the communication medium that is responsible for creating a bottleneck to effective communications. In some instances, it is one or more elements of the computing apparatus linked by the communication media that causes a bottleneck. For example, some commands instructing computing apparatus to perform certain tasks may be very short, but the tasks requested may be very complex and time-consuming. Under such circumstances, the command itself is so short that it will not result in congestion when transmitted over a communication medium, but, by instructing the computing apparatus to perform complex tasks, it may overload the capability of the computing apparatus to process other tasks simultaneously. Thus, a technique is also needed to manage the load placed on the computing apparatus by multiple commands that may come from multiple sources.
One example of a type of computing apparatus that can easily be instructed to perform complex tasks as a result of simple commands is a graphic display, for example, a video monitor or flat panel display. Such displays are comprised of large numbers of small image areas, the color and intensity of which may be individually selectively controlled. These small image areas are referred to as picture elements, or pixels. The display as a whole is a mosaic of pixels used to form characters, symbols, images, or other graphic elements being displayed. However, the process by which the pixels and, consequently, the image being displayed is changed requires time to effect the desired change of pixel colors and intensities. Thus, multiple commands instructing multiple simultaneous changes in an image may overload a graphic display device. Therefore, graphic display device can also be characterized as having a limited bandwidth, with the bandwidth expressed in, for example, pixels per second.
In the past, various attempts were made to accommodate data from multiple sources wherein the cumulative data rate of all such data to be transmitted exceeds the bandwidth of the medium over which the data is to be transmitted. One approach penalizes each of the multiple sources of data to an equal extent. This approach reduces the maximum data rate allocated to each of the sources of data. Thus, the rate at which data is transmitted is reduced for each source. For example, if the cumulative data rate at which the multiple sources would ideally send data is 20 megabits/second (Mbps), but the communication medium is limited to 10 Mbps of bandwidth, the maximum data rate allowed to each source could be cut by 50 percent to prevent overloading the 10 Mbps medium.
While this approach sounds simple and equitable, the effects of reducing the maximum data rate allowed by 50 percent will not affect all sources equally. For example, a source that is allocated, say 4 Mbps of maximum allowable data rate, might normally only transmit data at a rate around 1 Mbps, but occasionally transmit short bursts of data closer to the 4 Mbps rate. If its 4 Mbps allocation is reduced to 2 Mbps, the source will still be allocated twice as much bandwidth as it normally needs, so no adverse effects would be noticed while the data rate remains around 1 Mbps. Only during the bursts of data approaching the 4 Mbps data rate would the data transmission be affected by the 2 Mbps rate limit.
On the other hand, for a source that normally transmits data at a small, but relatively consistent, data rate, a reduction of data rate limit by 50 percent would severely affect the data transmission. Since only half the amount of data could be passed within the same amount of time, a backlog would instantly develop and continue to get worse as long as the lower data rate limit was in place and the consistent data rate attempted by the source was maintained.
Thus, a common reduction of all data rates in response to communication attempts in excess of the capacity of the medium does not fairly and evenly affect all data sources. Data sources exhibiting smaller and more consistent data rates are often affected more severely than data sources that have been allocated large data rate limits or that are more irregular in their data transmission.
Another approach uses a “first-come, first-served” principle, where bandwidth is awarded to the first requester in whatever amount the first requester desires. This approach obviously provides no disincentive for inefficient requests and unfairly penalizes later requesters.
Another scheme that has been used awards more bandwidth to sources exhibiting a high demand for bandwidth. Such an approach may consider the source requesting the most bandwidth to be the most important source and the other sources requesting less bandwidth to be less important sources. Thus, the scheme gives more bandwidth to what it deems to be the most important source, while penalizing what it deems to be the sources of lesser importance.
The problem with such a scheme is that sources attempting to be efficient and use as little bandwidth as possible ar
Bos Herbert
Butcher Lawrence
Hanko James
Northcutt Duane
Ruberg Alan T.
Nguyen Steven
O'Melveny & Myers LLP
Sun Microsystems Inc.
Vu Huy D.
LandOfFree
Method and management of communications over media of finite... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and management of communications over media of finite..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and management of communications over media of finite... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2913065