摘要:
A data system, such as a computer network or a telecommunications system, has a base station and a number of nodes called network interface units (NIUs). Each NIU is assigned to one of a number of groups of NIUs. The base station transmits special data to the NIUs indicating which groups of NIUs are currently authorized to transmit upstream messages to the base station, where only the authorized NIUs are allowed to transmit upstream messages. When message collisions in the upstream direction becomes a problem, the base station can reduce the number of NIU groups that are authorized at any given time to transmit upstream messages, in order to attempt to reduce the rate of message collisions. In one possible embodiment, each NIU is assigned horizontal and vertical group words that identify to which group it is assigned. The special data contain additional horizontal and vertical code words that are monitored by the NIUs. If the bits in the downstream code words appropriately match the corresponding bits in the NIU's assigned group words, then the NIU recognizes that it is authorized to transmit upstream messages to the base station. Otherwise, it is not. The horizontal and vertical code words correspond to two-dimensional patterns that indicate which subset of NIU groups are authorized at any given time. By cycling through a series of 2D patterns, the base station ensures that each NIU is authorized at least some of the time. Depending on the number of bits in the code words, a back-off scheme can be designed with different back-off levels, each of which authorizes a different number of groups to transmit. The base station can progressively decrease the rate of upstream transmission by selecting back-off levels with fewer and fewer NIU groups authorized to transmit.
摘要:
A method and apparatus for switching from an active to a standby controller in a redundant duplex carrier system without requiring retransmission of messages in progress during the switch. A main processor in the standby controller is periodically switched and made the active controller to ensure reliability of the system. The main processor in the active controller stops sending messages for a period of time before switching control to the standby controller. The main processor in the standby controller delays initiating any new communications until any current messages have been completed from the active controller. The main processor of the standby controller then retrieves any unprocessed messages that may have been sent to the active controller just before or during the switch from the active controller to the standby controller. The delay of new messages and retrieval of any unprocessed messages ensures that no data will be lost during the switch, thus removing the need for retransmission.