Automatic Router Configuration | Solutions and Recommendations

Automatic configuration of the WR is performed using a 4-state machine including a start up state, a learning state, an operational state and a site down state as depicted in Figure 1.

Figure 1: A state machine for dynamic configuration of a wireless router

 

The WR configures the RF/IP topology in the startup state, and refines the topology in the learning state. In the operational state, the wireless router handles a full traffic load, and continues to check if it meets the operational thresholds. Scheduled or unscheduled maintenance leads the WR into site-down state.

1) Start up state: In the starting state, connectivity is first established between the WR and the wire-line routers in the network. The wire-line connectivity is then used to establish connectivity between WR and its WR neighbors. Subsequently, MPLS paths or other suitable virtual circuits or IP tunnels are established among the WR neighbors to facilitate inter-router communication. The WR uses wire-line connectivity to learn from its neighbors the RF topology in the neighborhood, and establish wireless-specific connectivity with its neighbors.

After establishing the wireless connectivity with its WR neighbors, a WR exchanges RF impact information, including some or all of the parameters mentioned earlier. By exchanging this information, and negotiating the various operation parameters, the WR is able to determine or estimate a set of operating parameters that will help maximize radio coverage, minimize interference, and aid in providing a seamless coverage from cell to cell with smooth handovers. If no coordination could be achieved between the wireless routers in a neighborhood, Operation, Administration and Maintenance (OAM) server is contacted for resolving the differences. The OAM server then performs the RF impact analysis and responds with the operational parameters for the new site and the neighboring sites. The OAM server may also re-identify neighbor sites after parameters are agreed to, and store them in the configuration and parameter tables in the routers. The routers transition to the Learning state as shown the figure. Now, a RF system and network has been established by activating the wireless routers.

In the start up state, chores of the WR are: i) identification of the neighbors of the WR and preparation neighbor list, ii) interference impact, coverage, and other parameter analysis, iii) configuration of LSPs with neighbors, and iv) exchange and negotiation of power and handoff parameters with neighbors. In this way, the WR automatically configures itself for operation in the wireless network. Once these operations are completed, it transitions to the learning state.

2) Learning State: In the learning state, the WR continues to analyze, exchange and negotiate parameters, in order to minimize interference in the wireless network and to ensure that all the operational thresholds are met. The WR transitions into this state from the start-up state when RF power is up, and from operation state when either operational parameters change, or the WR neighbors change, or the operational thresholds are not met. In the learning state, parameters are re-negotiated and re-estimated based on the information given during transition from the operational state. Once operational thresholds are met for a specific period of time, the WR transitions to the operational state.

3) Operational State: In the operational state, the WR continues to monitor its operational thresh-olds periodically or otherwise exchanges information with its WR neighbor to ensure maximum efficiency and minimum RF interference within the wireless network. If the operational thresholds are not met, the wireless router transitions from the operational state back to the learning state for detailed analysis and evaluation of the configuration parameters and reconfiguration, as required, so that the operational thresholds can be met.

Also, if any of the neighboring routers change, affecting the topology of the network, such as a neighboring router failure, or a new router is added to the wireless topology, the router transitions from the operational state to the learning state, to reconfigure itself to suit to the new topology. In addition, if any parameters are changed due to any requests from its neighbors, the WR transitions to the learning state, for analysis and evaluation of operation using the new parameters.

4) Site Down state: The wireless router may enter the site down state from the learning state or the operational state if it requires either scheduled or unscheduled maintenance. Upon power up, the router will again transition back to the start up state for reloading and reconfiguration of the operational parameters. In this way, the wireless routers automatically adjust and account for changing conditions in the network to optimize operation of the network.

Mechanisms to Dynamically Configure the Router | Solutions and Recommendations

The proposed WR has been designed for dynamic configuration of its operational parameters. Configuration parameters of a WR are typically related to the site and technology used. Site parameters may be classified as geo-location, network operation, service configuration, and antenna parameters. The technology specific parameters depend upon whether the CDMA or GSM is supported and include technology-specific site parameters, but may be broadly classified into coverage, spectrum, channel, interference, control, and threshold parameters. At a finer level of detail, site Id, number of sectors/beams, sector/beam ID, latitude and longitude, sector/beam location, maximum radius of influence are typical geo parameters. Similarly, network configuration parameters include network interfaces (e.g. Tl, SONET, T3, etc.), site capacity, and network capacity. In the service configuration, we may have the list of various services supported, and the related directory agent (DA) addresses. The antenna parameters listed on a per sector/beam basis include the antenna type, digitized pattern, horizontal/vertical beam widths, max gain, and mechanical and electrical down tilts.

In the technology parameters class, maximum RTD (round trip delay), PER (packet error rate), FER(frame error rate), and percentages of blocked calls, access failures, dropped calls constitute the threshold parameters subclass. The coverage parameters subclass includes environment (e.g. rural or urban), path loss margin, technology specific hardware losses and gains, RF coverage prediction models, and traffic distribution maps. The spectrum parameters subclass consists of channel bandwidth, channel mask, channel number range, and maximum transmit power per channel. The channel parameters include the number of channels in the range, air capacity/bandwidth, minimum channel spacing, frequency use, frequency grouping, and hopping sequences. The interference parameters include interference thresholds, power control thresholds, channelization and sequencing, channel scheduling algorithms, RF interference prediction models, traffic distribution maps, and adjacent channel interference threshold. The control parameters subclass includes access parameters, intra-technology and inter-technology handoff parameters, and timing parameters.

In the following subsection, we present a very high level procedure by which a WR learns its parameters and configures itself in collaboration with its neighbors. The control logic and the operations performed in each high level state could be quite complex with several states for error paths and exceptions. For example, application of RF or IP discovery protocols in the startup state involves considerable information exchange between a WR and its neighbors.