MXPA98006650A - Medium access control (mac) protocol for wireless atm - Google Patents

Abstract

A protocol, method, and apparatus for managing network communications are disclosed which are particularly well suited for ATM communications across a wireless medium. Contiguous time slots within a frame are allocated to each node having traffic to send. Each node is assured a nominal bandwidth, and excess bandwidth is distributed by demand. The allocation of excess bandwidth can be dependent upon the size of the buffer at each node, as well as the time-criticality of each message. Nodes communicate their requests for allocation by appending such control information to the first of their transmitted packets. The allocation, of each node's transmit and receive time slots, is transmitted to all the nodes at the beginning of each frame. Thereafter, each node need not participate on the network until their allocated time periods, thereby allowing portable devices to enter inactive states to conserve power. The network is operated in a connection mode;connections are established in a relatively non-interfering manner by the use of periodically occuring beacons. Inactive, unconnected, nodes need only monitor the network during these beacon periods, further allowing for power conservation.

Description

ACCESS CONTROL PROTOCOL A. A MEDIUM (CAM) BY WIRELESS ASYNCHRONOUS TRANSFER MODE (MTA) Background of the Invention 1. Field of the Invention This invention relates in a general way to networks and communication protocols, with particular relevance in wireless networks, or other networks that require a minimum investment time. Specifically, it is aimed at a control protocol for the Wireless Asynchronous Transfer Mode (MTA), although the presented characteristics can be applied to alternative protocols as well. 2. Discussion of Related Art Currently, communication networks are formed by interconnecting devices by wires or cables, and each device having a protocol for sending messages along those wires and cables. In some cases, a portion of such a network can be implemented as a wireless connection, employing radio or infrared frequency signals between the nodes. Such wireless connections are point-to-point, they have a single communication device at each end, each tuned to another on a different frequency to that of the other devices in the same geographical area. A wireless network, on the other hand, is formed without physical connections between the devices, using, for example, radio frequency signals. Each device on the network is tuned to the same frequency, and each device conforms to a protocol to send messages on this common frequency. The protocol can allow communication between all the devices in the network or the protocol can restrict each device to communicate only with a master device. Wireless networks offer a significant logistical advantage over wired networks, by obviating the need for wires or cables running to each device. With the growth of the availability of multi-media technologies, and the growing demand for access to information, the potential residential or business-type market based on Local Area Networks (RAL) is growing. The ease of installation and expansion of a wireless network is certain to create a high demand for wireless RAL. For example, a central base station can provide wireless services, including voice, video and data, to all communication devices in a home, or a wireless base station can provide communication between all laptops in an office, or all computers on a campus. To be successful, however, the techniques and protocols employed in these wireless networks must not be significantly inferior to their equivalent wired networks, and they must be capable of providing compatibility with existing networks, particularly in terms of compatible service and administration criteria. Since the past decades, protocols have been developed to effectively and efficiently manage the transfer of information within the networks of communication equipment. A premise underlying the development of these network protocols has been a wired network infrastructure. In a wireless network, the assumption on which wired network protocols were developed may no longer be valid. Although most of the existing protocols are functionally extensible to wireless networks, their effectiveness and efficiency may be adversely affected by the lack of a direct connection between the devices. Additionally, associated with a wireless network is the probability of the presence of portable devices, powered by batteries or batteries. Although some wireless devices, such as desktop computers, televisions, and home theater systems will be powered by AC (Alternating Current) power lines, a number of wireless devices, such as phones, cameras and laptops will be powered by batteries. When providing a wireless network protocol, an architecture that allows energy conservation must be considered as well. Such considerations are rarely given to wired networks. A common protocol used for data communications in a wired network is the Asynchronous Transfer Mode (MTA). The MTA has been developed to process data at high speed, with different data rates, different quality of service (CdS) requirements (for example, data reliability considerations, delay, etc.), and different connection or disconnection paradigms. of communications by multiple means. The MTA is also suitable for multiplexing video, audio and data in the same medium, through the appropriate choice of parameters required for each one. Audio data, for example, does not require that the reliability of data packet errors be re-committed, but they do not tolerate excessive delay. Video data can generally suffer more delay than audio, but it is intolerant to delay the fluctuation. These considerations of delay and error rate of the packet are better supported by a connection oriented to the service, where the parameters are negotiated and established at the beginning of each connection. For optimal performance, the MTA adopted an end-to-end error detection method, assuming the premise that the percentage of errors associated with the communication medium, such as a wired fiber-optic medium, was minimal. Only the terminal equipment verifies the errors; if an error is detected, a retransmission request is sent to the originating transmitter. The MTA also blocks any services for which the required CdS can not be guaranteed. These features, and other facets of the MTA, are very effective for multi-media communication over a wired network, but they are the factors that are contrary to the characteristics of existing wireless networks. The existing efforts to build a wireless local area network (RAL) are focused on emerging standards, such as IEEE 802.11 in the United States, and HIPERLAN in Europe. These standards do not take into account quality of service (CdS) requirements based on the MTA for data traffic and real time, particularly in the area of delay. In a typical wireless network, delays will increase exponentially in an overloaded network, when each transmitter contends for access. Most wireless networks operate using some form of collision detection and a reception acknowledgment protocol. For example, each transmitter will listen during a period of silence, and then transmit a packet. If another transmitter does not transmit at the same time, the receiver will receive the packet and will recognize the reception to the transmitter. If another transmitter transmits simultaneously, however, a collision occurs, the intended recipient receives mutilated or unintelligible messages and does not recognize the sending. After the acknowledgment of a failed reception, each transmitter will try to transmit again, hoping not to do it simultaneously. Wireless networks are characterized, in general, by having high percentages of errors, unpredictable delays, and requiring detection and intermediate correction of errors. In this way, typical wireless networks are inherently unsuitable for traffic in the MTA. Clearly, an important aspect in the design of a wireless MTA is that the control protocol that specifies the method of access between multiple users to the same medium, the protocol of Access Control to the Medium (CAM), must satisfy the basic requirements of the MTA, particularly in the area of delay considerations. One such protocol is the Assignment protocol of Dynamic Time Interval (ATD ++) designed to be used within a European project, Mobile Broadband System (SBM). This protocol was based, however, on the assumption that the communication paths in an uplink and a downlink, to and from a base station, were each performed at different frequencies. This reduced the investment time for control and recognition, but required that all stations accommodate transmission and reception on two discrete frequencies and without interference. Thus, it appears that the transformation of a wired network protocol, such as the MTA, to a wireless network protocol typically requires additional time or frequency or both. The purpose of this invention is to minimize the time required to communicate information within a network, without requiring additional frequency allocation, which supports both the MTA CdS concepts. Although the presented invention is particularly applicable to wireless networks in the MTA, the embodied principles are equally applicable to minimize the time required to transmit information over other wired or wireless networks as well.

BRIEF DESCRIPTION OF THE INVENTION Essentially, the invention describes a method for controlling access to a communication medium within a wireless network protocol, consistent with a network structure and protocol in the underlying MTA. The access control according to this invention comprises the following characteristics: - a structure of Control Data Block (BDC) to minimize investment time; - a Superinterval data structure for the signaling of an Ascending Link; a generalized Interval Assignment Policy for the CAM level, which integrates the allocation of intervals for the uplink and the downlink; - a Separator Overflow Control Policy at CAM level; and, an implementation based on the Guidance Signals for CAM administration services. The optimized Control Data Block (BDC) minimizes the investment time associated with transmissions in a wireless network, consolidating control and downlink information, and optimizing transmissions of control information and uplink of each transmitter . In addition to minimizing the investment time, this structure also provides a determinant nature to the communication patterns, allowing the terminal devices to perform other functions during periods of inactivity, or, in the case of devices powered by batteries or batteries, allowing them to conserve energy during those periods of inactivity. Associated with the Control Data Block is the use of a Superinterval architecture, to allow the incorporation of the uplink control information, in a data transmission through the uplink to optimize the communications of each of the devices wireless to a central system or base. This structure allows a minimum expenditure, particularly the expense required to synchronize the packets of each transmitter. The generalized interval allocation policy supports the MTA CdS concept of a nominal bandwidth allocation for each device, and, in conjunction with the Separator Overflow Control policy, allows reassignment of bandwidth for increments sudden packets, consistent with the goals of the MTA. Additionally, any remaining excess bandwidth is allocated efficiently between the devices to optimize overall performance and CdS factors. Consistent with MTA CdS concepts, each device in the wireless network negotiates a minimum level of service, with the understanding that, sometimes, the device will require more service, and other times, less. The MTA server, in this negotiation, promises to provide the capacity to support the nominal load, as well as the sudden increases of heavier load. The Overflow Control of the Separator, in accordance with this invention, performs two functions related to this concept of the basic MTA. It gives Priority to the transmissions to avoid the overflow of the separator during the periods of sudden increase, and at the same time, as necessary, it imposes the concepts of CdS of the MTA penalizing those devices that do not conform to those concepts. The implementation based on Guidance Signals for the administration of CAM services was optimized to reduce the synchronization time required between the devices, as well as to allow the inactive devices to have access to the network again when required. According to this invention, a short orientation signaling period is intermixed at relatively long but regular intervals in the data stream. Within these orientation signals, synchronization and control signals will be found to provide the highest level of synchronization and administration to the wireless network. By providing this higher level of synchronization and control, the cost of synchronization and control within each packet can be minimized, thereby improving the overall efficiency of the network.

BRIEF DESCRIPTION OF THE DRAWINGS FIGURE 1 shows a network of wireless devices. FIGURE 2 shows a Stratified Protocol model for wireline and wireless communications in the MTA. FIGURE 3 shows the structure of the Control Data Block. FIGURE 4 shows a flow diagram for the interval assignment. FIGURE 5 shows a flowchart for processing the laxity tail. FIGURE ß shows a Guidance Signaling method.

Detailed description of the preferred embodiments of the invention Figure 1 shows a wireless network, comprised of wireless terminals, IT, base stations, EB, and hybrid stations called mobile nodes, NM. To complete, the connection of the wireless network to a wired network of the MTA 101 is also shown through a gateway G. The figures la and Ib show a wireless network with a centralized architecture, or base station. The figures le and ld show a wireless network with a distributed or ad hoc architecture. In FIG. 1, the base stations EB are connected to a CC switching center in a tree topology. Wired Network packets from MTA 101 travel through gate C, and are routed to the appropriate base station EB via the CC switching center. The base station EB transmits and receives packets to and from the appropriate IT wireless terminal as will be discussed below. In Fig. Ib, the base stations EB are connected in an annular topology. The packet passes from base station to base station, until it arrives at its destination base station. The destination base station removes the ring packet and transmits it to the appropriate wireless terminal. In Figure 1, each NM wireless device can operate as a base station as well, so that its transmission can be received by all other wireless devices in the network. A transmission node topology is shown in Figure 1d, where node 120 is unable to communicate directly with node 122, but node 121 is provided with the ability to transmit packets from 120 to 122. Within each of those topologies, the invention will provide effective and efficient communication. To facilitate understanding, the topology of the base station of figure 1 will be assumed, but the implementation of the characteristics of this invention to other topologies would be evident to one skilled in the art. Figure 2 shows the stratified protocol model for the integration of the wired and wireless MTA via the Base Station. At one end of the communication path is a wired MTA device 210, such as the cable TV transmitter. At the other end of the communication path is a wireless MTA device 230, such as a portable television. Interconnecting those devices is the wired means 201, and the wireless means 202; the transformation from wired to wireless means takes place in the base station 220. This base station 220 could be, for example, a transmitter that is centrally located in a home, which is connected to the services provided by the MTA of, for example, a cable TV provider. Wired attributes specific to the communication path are contained in the FIS layer (Physics) Wired 221 and Wired CAM layer 222. Those protocol layers transform the physical signals in the wired medium into data packets conforming to the MTA at the MTA 223 level, which are independent of the physical medium employed. The wireless attributes specific to the communication path are contained in the wireless CAM layer 224 and the wireless F1S layer 225. These layers convert the packets that conform to the MTA into physical signals which can be transmitted over the wireless medium 202. The layers Similar wireless CAM 234 and FIS 235 are contained in the wireless terminal 230 and are intended to convert those physical signals into data packets conforming to the MTA at 233. The capabilities of the wireless CAM and FIS 235 layers in the wireless terminal 230 they can be significantly smaller than those provided by the base station 220, but the main function that they perform, of converting the physical signals to signals that conform to the MTA, is identical. The structure of the Control Data Block (BDC) for communication between the base station and the wireless terminals is shown in Figure 3. The Data Block of. Control, as its name implies, contains Control and Data. In accordance with this invention, the BDC is structured to minimize the investment time, in particular to provide a deterministic time for transmissions in this otherwise random process. Conventionally, in a wireless network, when a node has a packet, or set of packets, to transmit, it transmits each packet as long as it detects a period of silence. As noted above, this results in unpredictable delays and fluctuations in delays. In accordance with this invention, the base station will assign the available data ranges to the individual wireless terminals. This informs all the wireless assignment terminals in the signaling period of the Base Station 310 of figure 3. The base station transports this assignment, informing each wireless terminal when to start receiving the packets, how many packets to receive, when transmitting the packages, and how many packages to transmit. By doing this, each wireless terminal can subsequently allocate its time out of communication to other tasks, or, in the case of equipment powered by batteries or batteries, can set a timer to interrupt at the appropriate time, and enter an inactive state interim Although the BDC can be of variable length, each wireless terminal will also know when the next one will occur, BDC, knowing when the last packet will be transmitted in this BDC. Alternatively, to avoid very short DCB continuous transmissions, in the case that no packets are communicated, the signaling period 310 may contain an explicit notification of when the next BDC will occur. In this signaling period 310, the base station will also include maintenance operations and messages, as well as sequencing information for the currently inactive wireless terminals, as will be discussed with reference to period 340.

Consistent with traditional wireless communication,. each transmitter will be required to transmit a synchronization pattern to allow the intended receiver to establish a common synchronization base. The base station will transmit its synchronization pattern at the beginning of the signaling period 310. To avoid having to establish this synchronization with each wireless terminal receiver independently, when sending packets to each terminal, the base station will immediately follow the signaling period 310 with the transmission of all its packets in the downlink during the data period of the Downlink 320. Because all wireless terminals will have to synchronize with the base station to receive the signaling information, each wireless terminal that has been * programmed to receive the base station packets you simply need to maintain this synchronization until your scheduled reception time. Note that, in this preferred embodiment, the terminals powered by batteries or batteries may not be able to enter a completely inactive state, because if they do so, they may lose synchronization. Thus, consistent with this invention, the order in which the packets are transmitted to the terminal of the base station may depend on whether a particular terminal is limited by the power supply or not. In this way, packets for terminals powered by batteries or batteries will be transmitted from the base station first, allowing those terminals to enter an inactive state quickly. Alternatively, the resynchronization patterns could be transmitted with each set of packets to be sent to each transmitter. The data period of the downlink 320 comprises the transmission of all the programmed data packets from the base station to the wireless terminals. In accordance with this invention, each wireless terminal will receive all its scheduled data packets as a sequential set. By doing so, any required resynchronization within the first data packet can be achieved, and all subsequent data packets can contain a total data load, thereby increasing the overall efficiency of the packet information transfer. And, as stated above, this scheduled transmission of a set of sequential packets allows wireless terminals to allocate their time out of communication efficiently. The data period of the uplink 330 comprises the transmission of all the programmed data packets from the wireless terminals to the base station. As with the transmission of the base station, each terminal will transmit its assigned number of packets sequentially, starting at its assigned time. Unlike the transmissions of the base station, however, according to this invention, the first transmission interval of each terminal will be a Superinterval, which allows the transmission of a data packet, as well as a control packet annexed to this data package. By assigning this larger initial time interval, instead of inserting the control information into a data packet, higher efficiencies are obtained. If the control information were contained in a data package, it would be necessary to assign a complete package for the control, or it would be necessary that the original packages were modified to contain this information. Because the first packet of each transmitter is easily identified, having been programmed by the base station as discussed above, it is a significantly simpler matter to extract the attached control information, instead of decoding one or more data packets for extract the information included. The control information that is transmitted from each wireless terminal comprises the information required by the base station to administer the network and assign the bandwidth. This will contain the number of packets that the wireless terminals have reserved to transmit, and according to another aspect of this invention, its priority, as will be discussed later. This should also contain the operations and maintenance information, as would be typical of traditional network control protocols. The Alert period 340 is used by the currently inactive terminals to request a state of inactivity. Depending on the protocol established, this can take many forms. If the need for immediate attention is not imperative, a voting technique can be employed. For example, in signaling period 310, the central station could announce a terminal identifier, from among a list of currently inactive terminals. Subsequent BDC periods 310 may contain other terminals identified from this list, in a tournament form. If the identified terminal needs a state of inactivity, it will transmit a signal in the period 340. In the preferred mode, this signal could be minimal, to reduce the time allotted by such a vote. If the signal is present, the base station will assign one or more uplink intervals to this terminal at the next BDC. The identified terminal could subsequently transmit any control information in its assigned super-slot. Depending on the desired operating factors, the protocol could be set so that after receipt of an alert signal, the identified terminal receives its nominal bandwidth at the next BDC. Alternatively, the identified terminal may receive only one interval assignment, and the appropriate bandwidth could be allocated, depending on the number of specified packets that must be transmitted, according to the content of the control information appended to the aforementioned superinterval of this terminal. In the first scenario, excessive bandwidth may be assigned, in the second, excessive delay may be incurred. Also, alternatively, a full-time superinterval could be assigned during the Alert 340 period, to allow proper transmission of a packet and uplink control information. In the alert voting scenario, a network containing many inactive terminals may exhibit an excessive delay between the time of a terminal that has a packet to send and the time that is identified in the voting process. In accordance with this invention, alert period 340 may be structured differently to allow a more timely response, or better utilization of bandwidth, as would be apparent to those skilled in the art. For example, if a particular terminal were particularly sensitive to such delays, the voting sequence could be modified to vote that terminal more frequently than the others. 0, a method without voting can be used. For example, a containment method could be used within this period. The base station will have to be identified when this alert period 340 occurs, during the signaling period 310. Any terminal that needs to signal an alert request would simply transmit an identifier during this period, and the base station would allocate one or more intervals to this terminal in the following BDC, as detailed above. If a collision occurs, the identifier is not received, the absence of an assignment in the following BDC will notify. to the requesting terminal that must generate another alert request in the next warning period 340. Combinations of the voting and contention-based processes could also be used to request service, as would be evident to a person skilled in the art. As discussed, each active wireless terminal will notify the base station of the number of packets remaining after the current BDC transmission that must be transmitted to the base station. If this number is zero, the identifier of the terminal is placed in the list of the inactive terminal. If the number is not zero, the base station could assign intervals to this terminal in the next BDC. The base station must also allocate intervals for the received packets that need to be transmitted to the terminals, as well as intervals for the transmissions of all other active terminals. Consistent with the premise of the MTA discussed above, it is likely that at certain times of peak activity, there is insufficient bandwidth, ie an insufficient number of intervals within a BDC to accommodate the number of packets that need to be transmitted. Although the size of the BDC is variable, and as such could be extended to accommodate the transmission of all current packets, the size of the BDC should be restricted. The size of the BDC must be restricted to ensure that all terminals are served with sufficient frequency to avoid excessive delay, particularly terminals that conform to the allocation of bandwidth negotiated in exchange to ensure at least a nominal service. In the preferred mode, the BDC period would be limited to no more than one millisecond; With the present technology, this allows approximately 50-100 data packets to be transmitted in the MTA within each BDC. Those 50-100 data packets could eventually be assigned to the wireless terminals during the aforementioned negotiation phase, such as • each connection is made within the network. If the negotiation is not successful, for example if the base station does not have sufficient bandwidth, currently available to satisfy the minimum functional requirement for the terminal, the connection is not made. For example, voice communication typically requires a minimum bandwidth of 3KHz; if the base station does not currently have 3KHz available, a busy signal can be returned. Such negotiation scenarios are well known to those skilled in the art, and it is not pertinent to mention them in this invention except that, as stated above, each connected terminal must be guaranteed a CdS, which depends on a nominal bandwidth agreed. This nominal bandwidth is specified in terms of the number of packets, which, on average, will be communicated to and from the terminal in each period of maximum BDC. Thus, if the maximum BDC period is one millisecond, the bandwidth requirement of 15 kilopacks per second would be equal to the allocation of 15 nominal packets per BDC. Consistent with the premise of the MTA, however, the terminal to which this bandwidth of 15 kilo-per-second has been assigned can, for short periods of time, require a significantly greater bandwidth, and thus more your 15 packages per BDC assigned. In accordance with this invention, the base station will allocate intervals within each BDC to guarantee the nominal bandwidth, and allocate the unused slots to those terminals that require additional intervals in a prioritized manner. The prioritization, consistent with the premises of the MTA, will attempt to minimize the risk of data loss due to overflow of the separator, particularly for terminals with good behavior, and will attempt to provide an equitable allocation of any unused allocations. Because the BDC will contain uplink and downlink messages, the base station will handle the allocation of its address ranges with respect to the direction of the flow. By doing so, the unused downlink bandwidth can be allocated, as required, to transmissions on the uplink, and, correspondingly, unused uplink bandwidth can be allocated to transmissions on the uplink. downlink. In this and subsequent discussions, * no distinction is made between the downlink traffic of the base station and the uplink traffic of the wireless terminal. The allocation of the interval is effected by treating the downlink transmitter of the base station simply as another active terminal having traffic to send. For clarity, the term "node" will be used here later to encompass both the wireless terminals and the base station, with respect to the allocation of the uplink and downlink interval.

Figure 4 shows a flow diagram of an interval assignment policy according to this invention. As shown in 405, this assignment policy needs to be invoked only when the total of packages that are requested are communicated in this BDC is greater than the number of available intervals in the BDC. If the number of packets to be communicated is less than the number of available slots, all packets will be programmed to be communicated, at 410. If it is desired to communicate more packets than the number of available slots within a BDC, which packets should be determined they will be programmed in this BDC, and which ones will be postponed for a later BDC. This postponement can result in a loss of data, for example, if a node separator overflows, or there is an excessive delay. To minimize the average delay through the network, and satisfy the CdS criterion, all nodes are given at least their nominal bandwidth allocation, if required, at 420, as stated above, the base station will only establish a connection with a terminal if it has the nominal number of intervals available for that terminal in light of all other established connections and their nominal assignments. In this way, it is guaranteed that the assignment in block 420 will result in each node receiving its nominal allocation, as required. If, after the assignment of the nominal bandwidth to each active node, there are remaining intervals, those intervals could be assigned between the nodes with retentive data that should be sent. If all the nodes have infinite separators, a laxity factor could be used to determine a reasonable distribution of the remaining intervals to the nodes. Such a laxity factor, as will be discussed later, orders the packages by their need for relative transmission, immediately. A high laxity factor indicates a relative insensitivity to the delay; a low laxity factor indicates little tolerance to delays. The allocation according to this laxity factor, the packages with low laxity are assigned first, results in optimal performance and total clarity. But, as noted, such an allocation does not assume other consequences than the delay in its allocation. In fact, data loss can occur if the node has an insufficient buffer to accommodate data accumulation while waiting for its allocation. For example, the transfer of a file might be relatively insensitive to the delay, but at each node causing the delay, there must be enough memory to store the packets that continue arriving during this delay period. Thus, in practice, both the delay and the risk of data loss in the allocation process must be considered. In accordance with this invention, a portion of the remaining intervals are allocated to minimize the risk of packet loss due to the overflow of the separator, in block 430, and the remaining portions are allocated to maximize performance, while ensuring clarity in block 460. The size of the allocated portion to minimize the risk of data loss in 430 will be a function of the specific architecture and composition of the network. This proportion can be fixed, for example, half of the allocation for protection against overflow of the separator, half to maximize the yield. Or, it may be fixed initially, then modified by the base station or some other controller, depending on an evaluation of the overflow probability of the separator based on the current operation up to here. Block 430 is further detailed in Figure 4b. For each node that has a non-zero packet remnant to transmit, a separator overflow factor, fds (i), is calculated as the ratio of the amount of separator available, to the nominal assignment, in 433. The available separator is the size of the node separator minus the number of remaining packets to be transmitted. This overflow factor, in effect, is a measure of the number of nominal assignments remaining in the separator, or, equivalently, the number of BDCs remaining before the overflow of the separator, if the node separator is being filled at the nominal rate . As such, a smaller fds (i) indicates a greater probability that an overflow of the separator will occur. This factor of overflow of the separator is modified, in an additional way, to favor those nodes that conform to the premises of the MTA. One premise of the MTA is that the node will contain enough spacers to accommodate a dynamic allocation of bandwidth. The separator overflow factor, however, is a measure of how quickly the overflow of the separator will occur, giving higher priority to the nodes with smaller separators. A second factor, the size factor of the separator, fts (i), is calculated at 434. The size factor of the separator is a modified ratio of each size of the node separator compared to the average size of the separator in the network. If a separator has a separator size greater than the average, the ratio is set to 1, which will not result in the modification of the separator overflow factor. If the node has a lower average separator size, the ratio will be less than 1, and this relationship will reduce the priority given to this lower average node, as will be shown in 436.

Another premise of the MTA is that the node will often require less than its nominal allocation, to counteract those times when it requires more. A third factor, the behavior factor, com (i), is calculated at 435. This factor is the percentage of times that the node requires less allocation than its nominal. This is a measure of the low utilization of a node's allocation. This relationship will affect the priority given to the node, in direct proportion to its degree of low utilization, at 436. At 436, a total priority factor, the factor of overflow management of the separator, FMDS (i), is calculated. As noted above, the priority is inversely proportional to the overflow factor of the separator fds (i), and is directly affected by the size factor of the separator, fts (i) and the behavior factor, com (i). The greater the resulting FMDS (i), the greater the priority that must be given to this node. This factor is used in 442 to allocate the bandwidth to each node to avoid overflow of the separator while giving preference to the nodes that conform to the premises of the MTA. In 437, the sum of FMDS factors is accumulated. After calculating the FMDS for each node that has unallocated remnant requests, shown by the circuit formed by 432-438, the portion of the Overload Assignment amount, AS, assigned to each of those nodes is calculated in the circuit formed by 441-443. Each node is given an allocation in direct proportion to its calculated FMDS factor, in relation to the total of all FMDS factors calculated, at 442, up to its requested quantity. This allocation for each node is returned at 449, and will be added to that nominal allocation of the node as discussed above, and that node's laxity assignment as will be discussed later. Consistent with this invention, the FMDS factor can be modified in numerous ways. For example, a higher threshold can be defined for all or each of the overflow factors of the node separator. If the overflow factor of the node separator exceeds this threshold, its FMDS could be set to zero, thus avoiding any overflow allocation for the nodes that show little risk of overflow. Similarly, the FMDS (i) can be restricted to be less than some maximum number, to prevent any node from monopolizing the assignment when its separator overflow factor approaches zero. All unallocated remaining intervals are assigned according to the priority of a laxity factor at 460. As will be discussed later, each node maintains a packet queue, classified by its laxity values. Laxity is a measure of the delay in which each package can be incurred in an acceptable manner. A low laxity value indicates that the packet can not be delayed too long; a high laxity factor indicates a tolerance relative to the delay. Each node communicates the number of packets in its queue within discrete ranges of laxity values. For example, if three ranges are defined as low, medium and high, a node can inform the base station that it has 15 packages with a low laxity, 18 packages with medium laxity and 6 packages with a high laxity. This information is communicated to the base station during the assignment of the superintervals of the terminal, as part of the uplink control information mentioned above. When the intervals are assigned to each node, during the nominal assignment or the overflow assignment of the separator discussed above, the number of remaining packets in each node will be updated, in order of priority according to the laxity. If the node used for the previous example is assigned 2 intervals during the nominal assignment, and 8 intervals in the assignment of overflow of the separator, for a total of 20 intervals. Within this assignment the node will send the 15 packages of low laxity and 5 of the packages of average laxity. The node will be left with no packages of low laxity, 13 of medium laxity and 6 packages of high laxity. Any remaining unassigned intervals are assigned, first, to the low laxity packs, then to the medium packs and then to the high packs. In this way, if another node had any remaining low laxity packets, it would receive the assignment for all its low laxity packets, until the number of remaining intervals is assigned. This process is detailed in figure 4c. The circuit formed by 461-464 assigns the intervals in order of priority, in 463, until there is an insufficient number of available intervals to assign all the packages that have this level of laxity, in 462. If there are insufficient intervals for the assignment to all packages of equal level of laxity, the assignment in 466 will be distributed according to an allocation per tournament, to facilitate the calculation and administration of the queue, or according to a random assignment, for equity. The resulting allocations for each node are returned in 469. The total allocation for each node (the sum of the nominal assignments + overflow + laxity) will be communicated to the nodes at the beginning of the next BDC period, as discussed above. Figure 5 details the allocation and use of laxity factors in the individual nodes. Each node will place the received packets for later transmission in a Laxity Queue. At each node, a laxity factor is assigned when a connection is established. This laxity factor is a relative measure, and is assigned depending on the delay and loss of sensitivity associated with the traffic that will be communicated via this connection. Channels that are insensitive to delay are assigned a finite value, too, to avoid excessive delays. For example, a file transfer communication channel could assign a high laxity factor to each node in the connection. A voice communications channel, on the other hand, could assign a low laxity factor to each node. The exact assignment of the values of this value is not significant; The significant aspect is that the value must be consistent and comparable. In the preferred embodiment, the laxity factor is equal to the maximum delay, in units of time, permissible at each node along a connection, so that the sum of those numbers would be the maximum end-to-end delay permissible for communications acceptable In each node, the specified laxity factor is assigned to each packet when it is up. The insertion of the package in the laxity queue will be in the order of the assigned laxity factor of the package. If multiple packages have the same laxity factor, they can be placed in an additional order because of their sensitivity to data loss or other factors. For example, data packets could be placed above the voice packets with the same laxity factor.

In each BDC period, the node receives its transmission assignment, T, from the base station at 510 in Figure 5a. The T packages are removed from the queue, the lower laxity first, and transferred to the transmission queue for transmission during the BDC period, at 520. If any package remains, 525, its laxity factors are reduced by the time since the last BDC. The elapsed time is calculated at 530, and the circuit formed by 540-542 effects the reduction of each packet in the laxity queue. In this way, the laxity factor of each packet is an indication of the time remaining before the subsequent delays become problematic. For effective allocation at the base station, wireless terminals must communicate the number of packets that will be transmitted at each level of laxity. Therefore, the base station can allocate more intervals to nodes that have packets with low laxity levels than to nodes that have high laxity levels. This information will be transmitted to the base station via the control information that is attached to the first packet of each terminal, as discussed above. However, the laxity factor can cover a large interval, and transmitting the number of packets in each of the possible values, could be a process that consumes a lot of time and bandwidth. Therefore, according to this invention, the priority factors are quantified in a much smaller range of values. This quantification can be linear, for example, by dividing the interval of the priority factor into Q equal parts. The total number of packets contained in each of those Q segments of the priority scale would then be transmitted to the base station. According to this invention, the preferred embodiment comprises a non-linear segmentation of the priority factor interval, according to a logarithmic scale. By doing this, packets with low laxity values can be 'distinguished more finely in allocation processes, than those with larger values, and immediately less important. Figure 5b shows the quantification according to this invention. The NumPac is an array of values, the NumPac (q) contains an account of the number of packets in each of the Q possible levels of laxity. Tables 560 to 564 represent a circuit, which is executed for each packet in the laxity queue. The level of laxity quantified by each packet is determined by the calculations at 561 and 562, and the NumPac associated with this quantization level accumulates the number of packets with its resulting quantization level at 563. After this calculation for all packets in the laxity queue, the resulting accumulations in the NumPac array are returned, at 565, for subsequent communication to the base station via the control information and the terminal superinterval assignment, as discussed above. The base station performs an accumulation of the similar level of laxity for its downlink packets, and the laxity allocation is effected, as noted above, regardless of whether the traffic is related to the uplink or the downlink. To further optimize the efficiency of a wireless network in accordance with this invention, a synchronization and management capability based on guidance signals is provided, as shown in Figure 6. Periodically, at intervals much greater than the nominal BDC period , a time is allocated for a period of communications of guidance signals, 610. The purpose of this signaling period is to separate the long-term management functions from the short-term functions described above, such as interval assignment and signaling. of BDC. The orientation signal will contain, for example, a master synchronization signal. All terminals on the network will use this synchronization signal to roughly adjust their local synchronization signals, perform filtering matches, and other adjustments or maintenance tasks.

Subsequently, in each BDC, the signaling of the synchronization only needs to contain the signaling required to make finer adjustments. Dynamic filter parameters, collected during the orientation signaling period may, for example, be used directly to initialize such filters directly before reception of the BDC signal, which is provided for a finer adjustment. The period of orientation signaling is also the period in which the establishment of the connection and the CdS negotiations mentioned above take place, as well as the tearing down of the connection and other events with relatively long or infrequent deadlines. Placing these activities within this slower frequency synchronization minimizes the impact of such activities on the CdS of the established connections. Establishing the connection can, for example, require the communication of thousands of packets. If we force these communications to happen to a transfer of twenty packets in each orientation signal, and the orientation signals occur only every tenth of a second, it will take five seconds to establish the connection. The impact on the network, however, using this method, will be significantly less than if a thousand intervals were required that compete with the established connections in a period of the BDC, and the connection was made in less than one second, by assignments taken away from the established connections. The limitation of the effective bandwidth for the establishment of the connection lengthens the time required to establish each connection, but results in a better operation once the connection is established. Additionally, maintaining a separate scheme for connecting and disconnecting terminals also provides efficiency to the structure and operation of the BDC described above, in particular with respect to alert period 340. Absent this guidance signaling, the structure of the BDC and the protocol must accommodate the alert signaling of any potential wireless terminals. Such an extensive task can be difficult to accommodate * in the voting scenarios discussed above.

By segregation of the connection establishment of the BDC administration routine, the list of inactive terminals used for voting can be reduced to a minimum, so that only those currently connected terminals are also currently inactive. The periodic guidance signaling also allows more efficient energy conservation. A cordless phone currently disconnected, for example, could verify a period. of guidance signals, knowing that those connections were established only during such periods. If this was not ordered in the orientation signals, you could program your watch to self-interrupt just before the next scheduled guidance signal period, and enter an idle state for an interim period. To achieve the established advantages, it is preferred that the period of orientation signals occur at very regular intervals. As such, in light of the fact that the BDCs are of variable length, there may be groups in the BDC series, as shown in 620. Because the base station allocates the specific transmission time of each node, a BDC it can be easily distributed to avoid excessively long gaps or intervals, as shown in 631-633. Without the partition, the gap or interval 632 could extend from the beginning of the period of the BDC 631 to the beginning of the orientation signal period 611. Provided that such gap or interval is sufficiently long for the base station to transmit its signaling information of control (310 in Figure 3), the base station can program the time of the assigned transmissions to avoid the period of orientation signals. A portion of the BDC will be transmitted in 631, a smaller gap or interval will be obtained 632, and the rest of the BDC will be programmed by the base to occur at 633. Alternatively, some variation in the regularity of the orientation signals can be allowed, and this could be incurred immediately afterwards. of a BDC, or a portion of the BDC, thereby eliminating all gaps or intervals not used over time. The foregoing only illustrates the principles of the invention. It should be appreciated that those skilled in the art will be able to devise various arrangements which, although not explicitly described or shown here, incorporate the principles of the invention and are thus within their spirit and scope.

Claims (11)

CHAPTER CLAIMEDICATORÍO Having described the invention, it is considered as a novelty and, therefore, the content is claimed in the following: CLAIMS:

1. A communication network, characterized in that it comprises a plurality of nodes coupled to a central node, wherein an active node is one of such nodes that has one or more packets to be transmitted, the active node is arranged to communicate a request for time allocation within a period of the block of each active node, the central node is arranged to assign a transmission time and a duration of the transmission within the period of the block to each of the active nodes, depending on the assignment request, and for communicating such allocation of time and duration to each of the active nodes within the period of the block, the inactive node is arranged to transmit packets according to the allocation of time and duration within the period of the block.

2. The communication network according to claim 1, characterized in that some or all of the communication and transmission is via a wireless communication means.

3. The communication network according to claim 1, characterized in that the assignment of the time and transmission duration of each active node is determined based on one or more allocation parameters associated with each of the active nodes.

4. The communication network according to claim 3, characterized in that a nominal assignment is found associated with each of the active nodes, and wherein one of the allocation parameters is the nominal assignment associated with each of the active nodes. the active nodes. The communication network according to claim 3, characterized in that each of the active nodes comprises a separator for transmitting or receiving packets, and one of the allocation parameters is the size of the separator in each of the active nodes. The communication network according to claim 3, characterized in that the packets are MTA packets, and one or more of said allocation parameters comprise one or more Quality parameters of the MTA service. The communication network according to claim 1, characterized in that the assignment comprises the steps of determining a level of laxity associated with each of the packets, assigning the transmission time and duration of the transmission to each of the nodes assets, depending on the assignment request, and depending on the levels of laxity. 8. The communication network according to claim 1, characterized in that a level of laxity is associated with each packet, and, the assignment request of each active node comprises an account of the number of packets to be transmitted at each level of communication. laxity. The communication network according to claim 1, characterized in that the communication of the assignment request of each node comprises the steps of appointing the request to the first transmitted packet during the time assigned to the node in the current block period, or if * was not assigned to the node in the current block period, communicate an alert message after the last assigned time between the nodes in the current block period. 10. A communication device for communicating messages within a network, comprising a plurality of nodes, the device is characterized in that it comprises means for transmitting a request for allocation of time to transmit messages, and means for receiving an assigned first time and a first duration for receiving messages, means for receiving a second assigned time and a second duration for transmitting messages, means for receiving messages at the first time for the first duration, means for transmitting messages at the second time for the second duration. 11. A method for communicating packets within a network comprising a plurality of nodes, characterized in that the active node is one of the nodes that has one or more packets to be transmitted, the method comprises the steps of communicating a time allocation request within a period of the block of each active node, assigning a transmission time and a duration of transmission within the period of the block to each of the active nodes, depending on the allocation request, communicate the allocation of time and duration to each of the active nodes within the period of the block, and transmit the packets of each one of the active nodes according to the allocation of time and duration within the period of the block. SUMMARY OF THE INVENTION A protocol, method and apparatus for handling communications in a network is described, which are particularly suitable for communications in the Asynchronous Transfer Mode (MTA) through a wireless medium. Adjacent time slots are assigned within a block to each node that has traffic to send. Each node is assured a nominal bandwidth, and the excessive bandwidth is distributed according to demand. The allocation of excessive bandwidth may depend on the size of the separator in each node, as well as the critical nature of the time of each message. The nodes communicate their assignment requests by attaching such control information to the first of their transmitted packets. The assignment of each one of the time intervals of transmission and reception of the node is transmitted to all the nodes at the beginning of each block. Therefore, each of the nodes does not need to participate on the network until they are assigned time periods, thus allowing portable devices to enter inactive states to conserve energy. The network is operated in a connection mode; the connections are established in a relatively non-interference manner, by the use of orientation signals that occur periodically. The inactive nodes, which are not connected, only need to verify the network during these periods of orientation signals, also allowing the conservation of energy.