Channel selection method for distributed wireless ad hoc network, electronic device and storage medium

Abstract

The application discloses a channel selection method of a distributed wireless ad hoc network, an electronic device and a storage medium. The method comprises the following steps: obtaining a channel busy degree index by counting the ratio of the number of occupied time slots in a target channel to the total number of time slots of a superframe; determining the channel availability level of the target channel according to the channel busy degree index, a signal strength parameter and the number of upper network nodes that can be detected in the channel; for a first type of node configured to maintain the continuity of a backbone link in the ad hoc network, suppressing the first type of node from initiating channel switching; for a second type of node configured to bear an edge access or relay function in the ad hoc network, when a channel switching condition is met, sequentially traversing candidate channels according to a preset channel sequence, re-executing the evaluation of the channel availability level on each candidate channel, and switching a radio frequency operating frequency point to a target operating channel with the optimal channel availability level after re-evaluation; wherein the switching execution time of the radio frequency operating frequency point is aligned with the frame boundary of the superframe.

Description

Technical Field

[0001] This application relates to the field of wireless communication technology, and more specifically, to a channel selection method, electronic device, and storage medium for a distributed wireless ad hoc network. Background Technology

[0002] Wireless mesh networks are distributed wireless networks that require no fixed infrastructure and are autonomously discovered and dynamically constructed by nodes to form a mesh topology. They possess core features such as multi-hop relay, decentralized peer-to-peer interconnection, automatic routing, and self-healing. They are currently widely used in emergency communications, industrial sensing, low-altitude drone networking, and smart city sensing. Particularly in unlicensed millimeter-wave bands such as 24GHz / 60GHz, due to limited spectrum resources and an open electromagnetic environment, multiple nodes need to achieve synchronization, broadcasting, and service data transmission through time-division or contention mechanisms on a shared channel.

[0003] In typical real-world deployments, Mesh networks often face three types of coupled and complex operating conditions: First, a large number of fixed base stations and mobile nodes concurrently access and interact within the same coverage area, leading to fierce competition for channel resources; second, traffic radars, vehicle-mounted radars, 5G base stations, and other low-power devices operating in the same frequency band are distributed in urban environments, creating continuous or sudden external interference; and third, nodes are susceptible to the effects of buildings, vegetation, moving obstructions, and weather conditions during movement, resulting in non-stationary fluctuations in link quality.

[0004] To mitigate interference and congestion, existing solutions often employ passive frequency hopping strategies based on RSSI (Received Signal Strength Indicator) or packet error rate, or rely on a unified channel switching rule across the entire network. While these mechanisms can maintain basic connectivity in low-density, static interference scenarios, they still have significant limitations in high-density, dynamic network environments. Summary of the Invention

[0005] In view of the above, this application provides the following technical solution:

[0006] The first aspect of this application provides a channel selection method for a distributed wireless ad hoc network, including:

[0007] Using a superframe of preset duration as the statistical period, the ratio of the number of occupied time slots in the target channel to the total number of time slots in the superframe is calculated to obtain the channel busyness index.

[0008] Based on the channel busyness index, signal strength parameters, and the number of detectable upper-level network nodes in the channel, the channel availability level of the target channel is determined.

[0009] For the first type of node in an ad hoc network that is configured to maintain the continuity of the backbone link, suppress its channel switching;

[0010] For the second type of node in the self-organizing network that is configured to undertake edge access or relay functions, when the channel switching condition is met, the candidate channels are traversed sequentially according to the preset channel sequence, the channel availability level is re-evaluated on each candidate channel, and the radio frequency operating point is switched to the target operating channel with the optimal channel availability level after re-evaluation.

[0011] The switching of the radio frequency operating point is performed at a time that is aligned with the frame boundary of the superframe; whether the channel switching condition is met or not is related to the channel availability level of the target channel.

[0012] In one possible implementation, the first type of node includes a main gateway node and a main chain gateway node, and the second type of node includes a slave gateway node, a slave chain gateway node, and a dechain terminal node;

[0013] The specific method for suppressing the initiation of channel switching is as follows: the first type of node maintains its current working channel unchanged in order to ensure that the backbone topology does not oscillate.

[0014] In one possible implementation, the occupied time slot is determined by any of the following methods:

[0015] The system checks whether a valid preamble of a preset format exists within a time slot to be included in the network's service usage.

[0016] If the received signal strength reaches a preset occupancy threshold and no valid preamble is detected within the detection time slot, it is counted as external interference occupancy.

[0017] The channel busyness index is the sum of the proportions of various occupied time slots obtained statistically over multiple consecutive superframe periods.

[0018] In one possible implementation, for a Type II node in the initial network access phase, when traversing all candidate channels, the target working channel that simultaneously meets the following conditions is preferentially selected:

[0019] There is broadcasting from the upper-level network node, the channel busyness index is lower than the preset threshold, and the signal strength parameters meet the communication quality requirements;

[0020] If no channel meets all of the above conditions, a new network element will be established on an idle channel as a second-type node.

[0021] In one possible implementation, the channel switching conditions include:

[0022] The first scenario is when the channel busyness index of a network-connected node exceeds the preset busy threshold for multiple consecutive superframes, and is further classified as either service-induced busyness or interference-induced busyness: if the proportion of parseable service frames is higher, it is determined to be service-induced busyness, and the first type of node does not hop frequencies, while the second type of node hops frequencies; if the proportion of unparseable high-energy signals is higher, it is determined to be interference-induced busyness, and both the first type of node and the second type of node hop frequencies.

[0023] The second scenario is when the broadcast time slot cluster of an already connected node has a busy level of 0 and the upper-level network node is disconnected, which is determined to be a disconnected state.

[0024] The second type of node is allowed to perform frequency hopping search of existing networks according to the preset channel sequence.

[0025] In one possible implementation, the preset channel sequence is generated in any of the following ways:

[0026] A fixed traversal order based on ascending or descending order of the channel center frequency within the operating frequency band;

[0027] When persistent interference is detected, the channel list is divided into two groups, low-frequency band and high-frequency band, according to the interference frequency band, and channels outside the interference frequency band are selected alternately.

[0028] In one possible implementation, the superframe is divided into a broadcast time slot cluster and a time synchronization time slot cluster. The switching of the radio frequency operating frequency is completed within a predetermined transition period after the broadcast time slot cluster or the time synchronization time slot cluster ends, without interrupting the service transmission of the data transmission time slot cluster.

[0029] In one possible implementation, the time slot occupancy of the broadcast time slot cluster, time synchronization time slot cluster, and data interaction time slot cluster are independently statistically analyzed to obtain a busy index for each time slot cluster dimension. The busy index for each time slot cluster dimension is used to help determine the channel availability level.

[0030] A second aspect of this application provides an electronic device, including at least one processor and a memory connected to the processor, wherein:

[0031] The memory is used to store computer programs;

[0032] When the processor executes the computer program, it implements any of the above-mentioned channel selection methods for distributed wireless ad hoc networks.

[0033] A third aspect of this application provides a computer storage medium storing a computer program that, when executed by a processor, implements a channel selection method for any of the above-described distributed wireless ad hoc networks.

[0034] As can be seen from the above technical solutions, this application discloses a channel selection method, electronic equipment, and storage medium for distributed wireless ad hoc networks. The solution replaces instantaneous interference judgment with time-slot-level quantization evaluation, enabling nodes to distinguish between local network congestion and external interference. Through role-differentiated frequency hopping permission control, backbone stability is established as a protocol-level priority target, preventing edge disturbances from propagating into network-wide oscillations. Ordered channel selection during network entry and interference avoidance phases is achieved through sequence traversal and closed-loop reassessment. By anchoring physical handover to the superframe structure boundary, the impact on service transmission interruptions is reduced. Overall, the solution can simultaneously improve access success rate and service transmission continuity in low-altitude, high-density mesh and complex multi-interference scenarios, exhibiting high system robustness and engineering feasibility. Attached Figure Description

[0035] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of this application. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0036] Figure 1 This is a flowchart illustrating a channel selection method for a distributed wireless ad hoc network disclosed in an embodiment of this application;

[0037] Figure 2 This is a schematic diagram of the superframe composition disclosed in the embodiments of this application;

[0038] Figure 3 This application provides a schematic diagram of the evolution of an n-level chain gateway ad hoc network.

[0039] Figure 4 This is a schematic diagram of the structure of an electronic device disclosed in an embodiment of this application. Detailed Implementation

[0040] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0041] Figure 1 This is a flowchart illustrating a channel selection method for a distributed wireless ad hoc network disclosed in an embodiment of this application. See also... Figure 1 As shown, the channel selection method for distributed wireless ad hoc networks may include:

[0042] Step 101: Using a superframe of preset duration as the statistical period, calculate the ratio of the number of occupied time slots in the target channel to the total number of time slots in the superframe to obtain the channel busyness index.

[0043] The distributed ad hoc network described in this embodiment can be, but is not limited to, a mesh ad hoc network. After powering on and completing initial synchronization, the ad hoc network nodes first lock onto the currently residing channel as the target channel. The system uses fixed-length superframes as time scheduling units, that is, it uses fixed-duration superframes as the cycle period. The superframes are further divided into several equal-length minimum time slot units.

[0044] Figure 2 This is a schematic diagram illustrating the composition of a superframe as disclosed in an embodiment of this application. It can be combined with... Figure 2 Understand the components of a superframe. In implementation, channel occupancy can be statistically analyzed periodically using superframes, defining channel busyness as the ratio of the number of occupied time slots to the total number of time slots in the superframe. Specifically, nodes monitor the radio frequency environment in each time slot, recording the number of time slots with valid communication activity, and accumulating the statistics over multiple consecutive superframe periods. Channel busyness reflects the true time occupancy of the channel better than instantaneous received signal strength, effectively avoiding misjudgments caused by sudden pulse interference.

[0045] For example, in a 24GHz low-altitude sensing network, the superframe length can be configured to 20ms, containing 200 time slots of 100μs each. A node continuously counts 5 superframes, totaling 1000 time slots. If valid signal activity is detected in 320 of these time slots, the busyness index is 32%. This value provides the basic quantitative input for subsequent multi-dimensional evaluation.

[0046] Step 102: Determine the channel availability level of the target channel based on the channel busyness index, signal strength parameters, and the number of detectable upper-level network nodes in the channel.

[0047] After determining the busy level, the nodes synchronously collect the received signal strength and signal-to-noise ratio at the physical layer as signal quality parameters. During the broadcast monitoring period, they parse the network identifier field carried in the air frames and count the number of unique senders to reflect the detectable scale of the upper-level network nodes, i.e., a quantitative representation of gateway coverage density. The system inputs the busy level index, signal quality parameters, and the number of detectable nodes into a preset mapping rule, outputting discrete or continuous channel availability levels to achieve a comprehensive evaluation of the channel's "existence, load, and link quality."

[0048] That is, the node device determines whether the time slot is occupied by detecting whether there is a valid preamble or whether the RSSI reaches a preset threshold, and completes the statistics in multiple consecutive superframes to form a stable channel busy index; combined with RSSI strength and signal quality, it outputs the channel availability level.

[0049] It should be noted that the result of determining the channel availability level of the target channel is mainly used to evaluate the retention value of the currently occupied channel, and serves as one of the inputs for determining whether the second type of node enters the frequency hopping process. In the candidate channel traversal stage, the node re-performs the same set of evaluations on each candidate channel to obtain a new level, and compares it with the current channel level to finally select the optimal target working channel.

[0050] For example, a channel with a busy level of 40%, a received signal level of -68dBm, and the ability to detect broadcasts from two upstream gateways can be rated as "good"; another channel with a busy level of 75%, a received signal level of -55dBm, but no upstream network identifier is identified as a channel with strong external interference, and its availability level is "poor".

[0051] Step 103: For the first type of node in the ad hoc network that is configured to maintain the continuity of the backbone link, suppress its channel switching.

[0052] During the network entry process, self-organizing network nodes acquire their roles through an identity contention mechanism. In this embodiment, devices configured as Type I nodes assume backbone link functions, such as network main gateways or main chain gateways, and are responsible for the network's clock reference and routing relay. To ensure topology stability, even if such nodes detect increased channel congestion, they are logically suppressed from initiating channel switching requests, and are only allowed limited recovery in extreme anomalies, thereby ensuring that lower-level nodes can always maintain synchronization with them on a known fixed channel.

[0053] Step 104: For the second type of node in the self-organizing network that is configured to undertake edge access or relay functions, when the channel switching condition is met, the candidate channels are traversed sequentially according to the preset channel sequence. The channel availability level is re-evaluated on each candidate channel, and the radio frequency operating point is switched to the target operating channel with the optimal channel availability level after re-evaluation.

[0054] Whether the channel switching condition is met or not is related to the channel availability level of the target channel.

[0055] The second type of node includes gateways, chain gateways, and off-chain terminals that have not yet joined the network. When a node meets the switching conditions (such as exceeding the busy limit or being disconnected from the network), it no longer blindly and randomly hops frequencies. Instead, it loads a pre-set list of working channels, switches to candidate channels one by one in the order of the list, and stays for one statistical cycle to complete the re-evaluation of steps 101 to 102 mentioned above. It records the availability level of each candidate channel and finally selects the one with the best level as the target working channel for settlement. This closed-loop traversal mechanism avoids repeated frequency adjustment when multiple nodes hop frequencies concurrently.

[0056] The switching execution time of the radio frequency operating point is aligned with the frame boundary of the superframe. After completing the frequency hopping decision within the broadcast time slot cluster, the node does not directly issue a frequency hopping command, but suspends the channel switching request, waiting for the scheduler to capture the superframe frame boundary event. The switching action is constrained to be executed within the radio frequency idle window attached to the boundary. This window includes the guard interval after the end of the broadcast time slot cluster or time synchronization time slot cluster within the same superframe, as well as the reserved quiet period between the end of the superframe and the beginning of the next superframe broadcast time slot cluster, all of which belong to the non-service transmission segment near the superframe-level scheduling boundary. The frequency synthesizer only completes frequency reconfiguration and phase-locked loop stabilization at this time, ensuring that the entire frequency modulation process does not interrupt the normal service of the data interaction time slot cluster, realizing decision-making within the cluster and action within the boundary quiet window.

[0057] The channel selection method for distributed wireless ad hoc networks described in this embodiment replaces instantaneous interference judgment with time-slot-level quantization evaluation, enabling nodes to distinguish between local network congestion and external interference. By implementing role-based frequency hopping permission control, backbone stability is established as a protocol-level priority, preventing edge disturbances from propagating into network-wide oscillations. Ordered channel selection during network entry and interference avoidance phases is achieved through sequence traversal and closed-loop reassessment. Physical handover is anchored at the superframe structure boundary, reducing the impact on service transmission interruptions. Overall, the solution can simultaneously improve access success rate and service transmission continuity in low-altitude, high-density mesh and complex multi-interference scenarios, demonstrating high system robustness and engineering feasibility.

[0058] To facilitate a better understanding of the embodiments of this application, the following description uses Mesh networking as an example.

[0059] In a mesh network, hardware devices can be categorized into master gateways, slave gateways, main-chain gateways, slave-chain gateways, and off-chain gateways (or terminals awaiting network access) based on their identities. Generally, devices can acquire any of these five identities when entering the network. All devices can potentially assume one of these identities, and the network evolution depends on the protocol specifications. However, in some real-world scenarios, based on the device's specific business needs and hardware capabilities, devices can be pre-configured to participate in or possess only certain identities. This allows devices to find their place in the network more quickly, improving networking and communication efficiency.

[0060] The primary and secondary gateways can directly connect to the information platform within the network. Before joining the network, devices can determine whether they can directly communicate and interact with the information platform. If a device determines that it can connect to the information platform, it can, by default, compete for the status of primary and secondary gateways, and assume corresponding responsibilities and tasks within the MESH network.

[0061] The main chain gateway and the slave chain gateway cannot connect to the information platform directly. They need to rely on the MESH networking rules protocol to achieve mutual connectivity and transmit information to the master device (main gateway and main chain gateway), from which the master device forwards it to the information platform. Similarly, the slave gateway can also perform the same forwarding.

[0062] In the entire ad hoc network, the master devices (master gateway and main link gateway) are the primary maintainers of network connectivity in the current frequency band, while the slave devices (slave gateway and slave link gateway) can be considered as backups of the master devices at the same level in terms of topology logic. Under normal load conditions, the master devices independently complete the scheduling and information backhaul of the network elements under their jurisdiction, while the slave devices only undertake lightweight access-level relay. When the master devices are under resource pressure due to service backlog, the slave devices become more involved, undertaking the work of forwarding excess information and assisting in the maintenance of network operation, thereby achieving a balance between reducing backbone scheduling pressure and retaining expansion flexibility.

[0063] Chain gateway devices are divided into main chain gateways and slave chain gateways, and can be formed by multi-level connections. Data is forwarded level by level from the bottom n-level chain gateways to the main device, and then forwarded to the information platform.

[0064] Reference Figure 3 , Figure 3 This application provides a schematic diagram of the evolution of an n-level chain gateway ad hoc network, as shown in the embodiments of this application. Figure 3 As shown, Level 1 main chain gateway terminal 1, Level 1 slave chain gateway terminal 1, and Level 1 slave chain gateway terminal 2 can all receive broadcast messages from the main gateway and are all acting as Level 1 chain gateway devices. However, the Level 1 main chain gateway first obtains the identity of the main chain gateway, or its MAC address number is smaller than the MAC address numbers of the other two. Therefore, it competes to obtain the identity of the Level 1 main chain gateway, while the other two obtain the identity of the Level 1 slave chain gateway.

[0065] Similarly, the identities of terminal devices in Level 2 and Level 3 main chain gateways are also obtained according to this competitive specification.

[0066] In one implementation, the first type of node includes a main gateway node and a main chain gateway node, and the second type of node includes a slave gateway node, a slave chain gateway node, and a disconnected terminal node; the suppression of its initiated channel switching specifically means that the first type of node maintains its current working channel unchanged to ensure that the backbone topology does not oscillate.

[0067] In specific deployments, after a self-organizing node powers on, it first enters the identity negotiation window and listens for network broadcast frames within multiple consecutive superframes on the default scanning frequency. If a node hears a valid broadcast from a higher-level node and has sufficient downlink capacity, it backs off to request a lower-level role; if no superior declaration is found, it periodically sends the main gateway beacon with a higher transmit power, and after random backoff without conflict, it is established as the root node.

[0068] Once the roles are established, they are written to the non-volatile configuration area. For example, enumeration values ​​can be used to distinguish between the main gateway, main chain gateway, slave gateway, slave chain gateway, and off-chain terminal. The media access control layer reads this enumeration value during initialization. If it is determined to be a backbone node, a permanent suppression flag is set in the frequency hopping arbitration module. Thereafter, even if the routing layer generates a frequency hopping suggestion event due to service overload or energy increase, the underlying interface function will directly return execution failure without updating the radio frequency register. The physical layer will show that the channel is completely locked. If it is determined to be an edge node, only a temporary avoidance freeze window is set without long-term frequency locking.

[0069] Taking an urban low-altitude logistics network as an example, a main gateway is deployed on the roof of a tall building, and three main chain gateways are deployed on three streetlight poles to form a ring backbone. When thirty drones simultaneously transmit video during morning and evening rush hours, causing extreme congestion in the data cluster of a certain main chain gateway, in a traditional solution, the gateway might autonomously hop frequencies to find a clean frequency point. As a result, the entire network broadcast domain is instantly severed, and surrounding drones collectively lose synchronization and re-enter the network. However, in this embodiment, the main chain gateway would rather trigger service rate limiting or queue dropping than lose connection due to role inhibition logic. Flying nodes only sense rate fluctuations and do not lose connection, and the backbone clock anchor point always exists. At the same time, slave gateways and terminals can normally perform frequency hopping and traffic splitting.

[0070] This embodiment eliminates the incentive for indiscriminate frequency hopping across the entire network by explicitly defining the identities of five network elements and strongly binding frequency hopping permissions with topology functions. This makes the backbone channel behave as a logical constant during operation, significantly compressing the cause space of topology oscillations and shortening the convergence latency of the entire network in scenarios with large-scale node concurrency.

[0071] In one embodiment, the determination of occupied time slots may be as follows: detecting whether a valid preamble of a preset format exists within the time slot to be included in the local network service occupancy; or detecting that the received signal strength within the time slot reaches a preset occupancy threshold and no valid preamble is detected to be included in external interference occupancy. The channel busyness index is the sum of the proportions of various occupied time slots obtained statistically over multiple consecutive superframe periods.

[0072] In engineering implementation, two detection links can be driven in parallel at the beginning of each minimum time slot. The first link is preamble correlation detection. The baseband performs sliding correlation on the I / Q sampled stream after analog-to-digital conversion. The matching code is the synchronization preamble sequence agreed upon by this Mesh protocol. When the correlation peak amplitude exceeds the preset correlation threshold and falls within the time slot preamble window, it is marked as a time slot occupied with the identifier of this network or a compatible network. The second link is the energy detection path. The RF front-end turns on the receiver at the beginning of the time slot and integrates and averages to obtain the instantaneous received power. If the power value exceeds the preset occupancy threshold, for example, it is higher than the noise floor by 15 dB, even if the preamble is not hit, it is marked as having RF activity. It can usually be inferred to be an external system or radar-like signal.

[0073] The system can count occupancy as soon as it is satisfied on any path, while simultaneously accumulating two types of counts in the background: "local network occupancy" and "pure energy occupancy," providing a basis for subsequent interference attribution. To avoid fluctuations in busyness caused by occasional pulses, nodes maintain a sliding window with a length of several superframes. For example, the status of each time slot is continuously counted within five superframes, and the total number of marked occupied time slots within the window period is divided by the total number of time slots to obtain a smoothed busyness index. If necessary, hysteresis processing can be introduced, and the high load flag is only set when multiple consecutive windows exceed the standard. Taking a 24GHz traffic radar scenario as an example, the intersection radar transmits a broadband linear frequency modulated pulse every 100 milliseconds. If it happens to fall within a certain data time slot, without a preamble differentiation mechanism, the traditional solution would directly determine that the channel is unavailable and trigger frequency hopping. However, this embodiment identifies the characteristics of "no local network preamble and high pure energy ratio" through dual-mode judgment. The background can mark such channels as external intermittent interference rather than local network congestion, providing a basis for high-level decision-making.

[0074] In practice, whether a time slot being evaluated is occupied can be determined by monitoring whether there is a pre-set preamble within the time slot, or whether the RSSI value within the current frequency band reaches a threshold. These signal detection methods determine whether the current time slot is occupied. In a periodic superframe structure, the current channel is sampled and statistically analyzed: the number of time slots containing valid preambles, broadcast frames, and data frames within the channel is counted; the channel busyness index is calculated as (∑ number of occupied time slots) / (total number of time slots within the superframe period) × 100%; and combined with RSSI and signal quality, a channel availability level is determined.

[0075] This embodiment uses complementary determination of preamble and energy detection to enable the busyness index to have the ability to "identify the occupant". Combined with multi-superframe smoothing processing, it effectively filters out transient interference and hidden node effects, thereby improving the confidence and anti-spoofing of channel assessment.

[0076] In one embodiment, for a second type of node in the initial network access phase, when traversing all candidate channels, a target working channel that simultaneously meets the following conditions is selected: there is a broadcast from an upper-level network node, the channel busyness index is lower than a preset threshold, and the signal strength parameter meets the communication quality requirements; if there is no channel that simultaneously meets the above conditions, a new network element is established on an idle channel as a second type of node.

[0077] After powering on, the node enters scanning mode and tunes the radio frequency points sequentially according to the preset channel list index. Each time a frequency point is switched to, it silently listens for at least one complete superframe. It demodulates the broadcast time slot cluster frames to extract the network identifier, sender role, and transmit power. Synchronizing with the process described in the previous embodiment, it calculates the current channel busy level and infers the link margin based on the received signal level. The optimal filtering logic is hierarchical: first, frequencies without any valid network broadcasts are eliminated to avoid isolated inter-networks; second, frequencies with busy levels exceeding a preset safety threshold are eliminated to avoid high-load access; finally, frequencies with sufficient link margin to meet the demodulation threshold and minimal signal fluctuation are selected from the remaining set as the target channel.

[0078] If no suitable item is found during a full list traversal, the node will deselect a quiet frequency with the lowest energy noise floor and no broadcast activity. It will then periodically send self-built network element broadcast frames as a slave gateway or slave chain gateway to generate an independent micro-network identifier. Network element merging or roaming handover will be performed later when a stronger upper-level broadcast is detected. In the scenario of batch power-on of a drone swarm, nodes that complete the scan first occupy the backbone's high-quality channels. Later powered-on nodes, finding the original channels under high load, will automatically deselect a suboptimal idle channel to establish their own micro-network, achieving primary load balancing without centralized scheduling.

[0079] In this embodiment, when a node joins the network, it first performs a conditional channel screening, prioritizing frequencies with an upstream network, no congestion, and sufficient signal strength. This prevents new nodes from easily joining high-load or purely interference-prone channels, resulting in a higher success rate and better initial stability. If no suitable channel is found after scanning all channels, the node can settle for a quiet, idle channel and establish its local network. Later, when it flies to an area with formal network coverage or detects broadcasts from other main gateways, it can switch to the network according to the identity contention rules. This essentially provides nodes with a backup plan: "we can use it even without a network," eliminating the need for manual frequency band planning and allowing the network size to be adaptively adjusted.

[0080] In one embodiment, the channel switching conditions include:

[0081] The first scenario is when the channel busyness index of a network-connected node exceeds the preset busy threshold for multiple consecutive superframes, and is further classified as either service-induced busyness or interference-induced busyness: if the proportion of parseable service frames is higher, it is determined to be service-induced busyness, and the first type of node does not hop frequencies, while the second type of node hops frequencies; if the proportion of unparseable high-energy signals is higher, it is determined to be interference-induced busyness, and both the first type of node and the second type of node hop frequencies.

[0082] The second scenario is when the broadcast time slot cluster of an already connected node has a busy level of 0 and the upper-level network node is disconnected, which is determined to be a disconnected state.

[0083] The second type of node is allowed to perform frequency hopping search of existing networks according to the preset channel sequence.

[0084] The broadcast slot cluster busyness of 0, as described in this application, means that within a preset number of consecutive superframe periods, and within the broadcast slot cluster time period specified by the superframe structure, a node fails to demodulate any valid broadcast preamble carrying the network identifier; that is, the broadcast slot cluster occupancy count is zero. This state, combined with the detectable number of upper-level network nodes being 0, is used to distinguish between "local network service congestion" and "loss of upper-level coverage (upper-level network node disconnection)," serving as one of the criteria for triggering a link-free frequency hopping search.

[0085] In implementation, the system maintains two independent counters. The first counter counts the number of consecutive times the total busyness or data cluster busyness exceeds a preset threshold. When this threshold is reached (e.g., five times), it is considered a congestion trigger. At this point, the node considers its network resources exhausted and enters an ordered frequency hopping interference avoidance process. In this paper, "network" refers to a logical self-organizing network domain that the node has joined or is attempting to join, sharing the same network identifier and hierarchical identity system. "Network resources exhausted" means that, among the occupied time slots counted on the node's working channel, the proportion of time slots with the network preamble continuously exceeds a preset busy threshold, causing the remaining time slot resources of that channel to be insufficient to meet the needs of new access or service scheduling. This state differs from channel busyness caused by external interference from other systems, where time slot occupancy mainly originates from energy signals without the network identifier. The second counter counts the number of consecutive times within a superframe where the broadcast time slot cluster cannot be resolved to a valid superior preamble. If the received signal strength simultaneously drops sharply by more than a certain decibel compared to the historical average, it is set to a disconnect trigger. At this point, the node has most likely flown out of coverage or been blocked by a large obstruction.

[0086] The two types of triggers employ different strategies: Under congestion triggering, nodes continue sequentially traversing from the next index of the current channel, hoping to find an empty channel in an adjacent frequency within the same network; under off-link triggering, nodes reset the traversal pointer to the head of the list and appropriately extend the dwell time on each channel to improve the probability of capturing sparse broadcasts, and even periodically switch back to the original channel while hopping frequencies to test whether to reconnect to the original main network. In the off-link state, the gateway also adds roaming logic; if it finds a broadcast from another main gateway in the same domain, it can smoothly switch ownership instead of immediately declaring itself the new master, thus avoiding micronet fragmentation.

[0087] Frequency hopping is an important method for avoiding network congestion, resisting interference, and finding network entry points. Based on the above-mentioned busyness assessment values ​​and the specific application scenario requirements, the criteria for determining whether to switch operating frequency bands via frequency hopping can be formulated.

[0088] In implementation, whether frequency hopping is triggered can be quantitatively determined based on channel busyness indicators. Busyness statistics include two sources: first, resolvable protocol signals, which can be used to distinguish payload types such as broadcast frames, data frames, and time synchronization frames based on protocol parsing and statistical frame headers, and to calculate the proportion of their occupied time slots to the total number of superframe time slots; second, unresolvable interference signals, which are determined to be external radio frequency occupancy when the received signal strength indicator reaches a preset energy threshold but the local network identifier preamble cannot be decoded. Both types of signals are included in the channel occupancy status: the former reflects the local network service load, and the latter reflects external interference causing busyness. By separately calculating the proportion of resolvable frames and the proportion of high-energy unresolvable frames, nodes can distinguish between two scenarios: "high service volume leading to channel saturation" and "high interference leading to channel unavailability," providing a basis for subsequent frequency hopping decisions.

[0089] When the operating frequency band provides only a single channel, nodes can only alleviate congestion by improving the efficiency of co-channel coordination, and there is no prerequisite for implementing a frequency hopping mechanism. When there are multiple available channels within the frequency band, a frequency hopping decision process can be initiated. In principle, all devices have the physical possibility of frequency hopping, but the main gateway and main chain gateway, as the main devices, are usually set not to actively hop frequencies because they are already maintaining the current frequency band network, and frequency hopping may only be allowed in special scenarios.

[0090] When a master device triggers frequency hopping, it needs to distinguish between two types of busy conditions. If the busy condition is due to queue backlog or excessive number of resolvable service frames leading to resource exhaustion (service-related busyness), the master device can remain on its original channel, while slave gateways and chain gateways, as slave devices, can hop to other frequency bands to share the frequency band pressure and make fuller use of the entire frequency band resources. Among them, chain gateways with limited hardware capabilities and only focusing on end-point data uploads can directly search for idle frequency bands for independent access. If the master device determines that the channel is continuously occupied by high RSSI but unresolvable interference signals (interference-related busyness), it means that the current frequency band is not suitable for communication needs, and the master device should also actively hop frequencies to search for frequency bands with less or no interference. If the busy condition is determined to be due to regular service congestion with a high proportion of resolvable signals but no strong interference, the master device should remain on its original channel, and only slave devices should expand capacity locally according to load conditions.

[0091] When a device senses congestion or interference in the main frequency band, it can autonomously weigh whether to continue to stay on as an auxiliary main device for forwarding or to hop to a new frequency band network. This hierarchical traffic distribution strategy allows for a balance between backbone stability and frequency band utilization.

[0092] This embodiment distinguishes between two semantic categories: congestion escape and disconnection self-rescue. This makes the frequency hopping behavior conform to the real physical process of link evolution, taking into account both the energy consumption control and service reachability of mobile nodes.

[0093] In one embodiment, the preset channel sequence is generated in any of the following ways: a fixed traversal order arranged in ascending or descending order of the channel center frequency within the working frequency band; when persistent interference is detected, the channel list is divided into two groups, low-frequency band and high-frequency band, according to the interference frequency band, and channels outside the interference frequency band are selected alternately.

[0094] The generation of the preset channel sequence includes two basic forms. The first is the basic factory default mode, where nodes are directly arranged in ascending or descending order based on the center frequency of each channel within the operating frequency band, forming a fixed traversal order. In this mode, the protocol stack does not need to maintain an additional channel weight database, the code size is small, the configuration is simple, and all nodes naturally adopt monotonic traversal logic in the same direction, naturally avoiding disordered collisions when multiple nodes hop frequencies concurrently. For example, in a 24GHz sixteen-channel system, nodes are uniformly configured to explore channels upwards one by one starting from 24.00GHz. Even without any network management intervention, the frequency hopping behavior of the entire network still exhibits a predictable search path.

[0095] The second type is the anti-interference orchestration mode. When a node detects continuous radiation at a certain frequency point during energy scanning, and the energy at that location is significantly higher than the ambient noise level for multiple consecutive superframes, it is determined to have encountered local interference. At this time, the node automatically activates the anti-interference traversal logic: instead of simply traversing from beginning to end, it virtually divides the entire channel list into two groups, low-frequency band and high-frequency band, according to the interference center frequency. The traversal pointer prioritizes alternating between the outer edge of the low-frequency band and the outer edge of the high-frequency band, far away from the interference zone. For example, it first tries Ch2, then Ch14, then back to Ch4, then Ch12, so that the search path quickly crosses the interference channel in a "leapfrog" manner, avoiding repeatedly stopping on the interference channel and wasting time.

[0096] In implementation, nodes can maintain a unified list of working channels and select different frequency hopping strategies according to actual scenario requirements. After each frequency hopping channel switch, a new channel evaluation is performed to determine whether the conditions for network formation or network access are met.

[0097] For node devices that have never joined the network, the entire operating channel frequency band needs to be traversed, and frequency hopping should be performed in a fixed order to avoid random switching. The network status of all channels should be evaluated, and channels with an upstream gateway, low channel congestion, and good signal quality should be prioritized for access. If no suitable channel is available, an idle channel should be selected to establish a new network element.

[0098] In the event of channel interference, choosing random frequency hopping or splitting the channel list in two and hopping between the two sides of the interfering frequency band can better and faster find a suitable working channel and achieve a better anti-interference effect.

[0099] This embodiment switches between factory-set single-order hopping and interference zone binary hopping order as needed. According to the aforementioned rules, when a node needs to change its operating channel, it no longer randomly tries blindly, but instead tries sequentially according to a pre-arranged order: under normal circumstances, it searches along the frequency band, and when encountering interference, it hops back and forth outside the interference band. This mechanism achieves faster interference avoidance speed and a low collision probability when multiple nodes hop frequencies simultaneously.

[0100] In one embodiment, the superframe is divided into a broadcast time slot cluster and a time synchronization time slot cluster. The switching of the radio frequency operating frequency is completed within a predetermined transition period after the broadcast time slot cluster or the time synchronization time slot cluster ends, without interrupting the service transmission of the data transmission time slot cluster.

[0101] A typical superframe, from front to back, consists of a broadcast time slot cluster, a short guard interval, a time synchronization time slot cluster, and a short guard interval and data interaction time slot cluster. The broadcast time slot cluster carries network-wide announcements and neighbor discovery, while the synchronization cluster distributes time slot allocation tables. Both are low-speed control frames and tolerate slight delays. The predetermined transition period, i.e., the guard interval reserved by the protocol, is when the RF power amplifier shuts down the transmit link. Frequency hopping decisions can be completed and buffered within the broadcast time slot cluster. During the cluster frame transmission completion interruption, the target frequency point is written to the RF task queue, driving the frequency synthesizer to start tuning and phase-locking within the guard interval. The power amplifier is re-enabled before the start of the next cluster, ensuring that the high-bandwidth service flow of the entire data cluster is not interrupted. For a typical configuration of a 20ms superframe and a 200μs-level guard interval, the mainstream millimeter-wave RF phase-locking time can be fully accommodated, resulting in a zero-perceptible handover for upper-layer applications.

[0102] This embodiment achieves seamless hot switching of services without increasing hardware costs by strictly embedding physical frequency switching into the silent window reserved by the protocol, enabling frequency hopping to be enabled normally rather than being limited to use during the dormant period.

[0103] In one embodiment, the time slot occupancy of the broadcast time slot cluster, time synchronization time slot cluster, and data interaction time slot cluster are independently statistically analyzed to obtain a busy index for each time slot cluster dimension. The busy index for each time slot cluster dimension is used to assist in determining the channel availability level.

[0104] The baseband independently statistically analyzes the effective preamble hits within broadcast time slot clusters, time synchronization time slot clusters, and data interaction time slot clusters to obtain the occupancy ratio for each cluster dimension. By observing the occupancy combination characteristics of the three cluster types, attribution differentiation can be made: high data cluster occupancy and low broadcast time slot cluster occupancy usually reflect a heavy local network service load; high broadcast time slot cluster occupancy but missing local network identifiers are often caused by external devices transmitting pseudo-broadcast signals; empty broadcast time slot clusters and a sharp drop in received signal level tend to indicate link interruption due to obstruction. When assessing channel availability, differentiated weights can be assigned to different cluster dimensions. For example, a mild penalty can be applied to data cluster occupancy, while a heavier penalty can be applied to broadcast time slot clusters without local network identifiers, naturally causing the scoring model to avoid pseudo-broadcast interference channels. The network management side can also read cluster-level statistical information to intuitively determine the nature of interference.

[0105] This embodiment separately counts the occupancy of the three types of clusters, enabling nodes to no longer focus solely on the total load but to distinguish whether the busy channel is experiencing traffic or interference. This provides an intuitive and interpretable basis for subsequent network optimization and frequency adjustment.

[0106] For the foregoing method embodiments, in order to simplify the description, they are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, because according to this application, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0107] This application also discloses an electronic device, Figure 4 This is a schematic diagram of the structure of an electronic device disclosed in an embodiment of this application. See also... Figure 4 As shown, the electronic device 40 includes at least one processor 401, at least one memory 402 connected to the processor, and a bus 403; wherein the processor and the memory communicate with each other through the bus; the processor is used to call program instructions in the memory to execute the channel selection method of the distributed wireless ad hoc network described above.

[0108] The memory is used to store computer programs;

[0109] When the processor executes the computer program, it performs the following: using a superframe of preset duration as the statistical period, it calculates the ratio of the number of occupied time slots in the target channel to the total number of time slots in the superframe to obtain a channel busyness index; based on the channel busyness index, signal strength parameters, and the number of detectable upper-level network nodes in the channel, it determines the channel availability level of the target channel; for the first type of node in the ad hoc network configured to maintain the continuity of the backbone link, it suppresses its channel switching; for the second type of node in the ad hoc network configured to undertake edge access or relay functions, when the channel switching condition is met, it traverses the candidate channels sequentially according to a preset channel sequence, re-executes the channel availability level evaluation in each candidate channel, and switches the radio frequency operating point to the target operating channel with the optimal channel availability level after re-evaluation; wherein, the timing of the radio frequency operating point switching execution is aligned with the frame boundary of the superframe; whether the channel switching condition is met or not is related to the channel availability level of the target channel.

[0110] Taking a UAV-borne millimeter-wave communication module as an example, the hardware consists of a monolithic radio frequency (RF) microwave integrated circuit, a system-on-a-chip (SoC) with a hard real-time baseband, non-volatile flash memory, and a power management unit. The RF integrated circuit integrates a low-noise amplifier, a power amplifier, and a frequency synthesizer, handling physical layer transmission and reception and frequency tuning. The SoC runs media access control tasks, implementing identity contention, slot-level busy statistics, and frequency hopping arbitration logic in the firmware. The RF abstraction layer encapsulates the frequency synthesizer tuning interface, which is invoked within a secure window by a superframe slot interrupt. The same hardware platform can operate in backbone frequency-locked mode or edge frequency-hopping mode by simply switching the running role through a configuration file, achieving all the steps of the aforementioned solution without the need for a customized RF front-end.

[0111] The execution entity of any of the channel selection methods for distributed wireless ad hoc networks described in the above embodiments includes a processor and a memory. The functional modules of the execution methods in the above embodiments are all stored as program modules in the memory, and the processor executes the program modules stored in the memory to realize the corresponding functions.

[0112] The processor contains a kernel, which retrieves the corresponding program modules from memory. One or more kernels can be configured, and the processing of backtracking data can be achieved by adjusting kernel parameters.

[0113] The memory may include non-permanent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0114] In an exemplary embodiment, a computer-readable storage medium is also provided, which can be directly loaded into the internal memory of a computer, and contains software code. After being loaded and executed by the computer, the computer program can implement the steps shown in any embodiment of the channel selection method for the distributed wireless ad hoc network described above.

[0115] In an exemplary embodiment, a computer program product is also provided, which can be directly loaded into the internal memory of a computer and contains software code. After being loaded and executed by the computer, the computer program can implement the steps shown in any embodiment of the channel selection method for the distributed wireless ad hoc network described above.

[0116] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.

[0117] It should also be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0118] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented directly by hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0119] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A channel selection method for a distributed wireless ad hoc network, characterized in that, include: Using a superframe of preset duration as the statistical period, the ratio of the number of occupied time slots in the target channel to the total number of time slots in the superframe is calculated to obtain the channel busyness index. Based on the channel busyness index, signal strength parameters, and the number of detectable upper-level network nodes in the channel, the channel availability level of the target channel is determined. For the first type of node in an ad hoc network that is configured to maintain the continuity of the backbone link, suppress its channel switching; For the second type of node in the self-organizing network that is configured to undertake edge access or relay functions, when the channel switching condition is met, the candidate channels are traversed sequentially according to the preset channel sequence, the channel availability level is re-evaluated on each candidate channel, and the radio frequency operating point is switched to the target operating channel with the optimal channel availability level after re-evaluation. The switching of the radio frequency operating point is performed at a time that is aligned with the frame boundary of the superframe; whether the channel switching condition is met or not is related to the channel availability level of the target channel.

2. The channel selection method for distributed wireless ad hoc networks according to claim 1, characterized in that, The first type of node includes main gateway nodes and main chain gateway nodes, and the second type of node includes slave gateway nodes, slave chain gateway nodes, and off-chain terminal nodes; The specific method for suppressing the initiation of channel switching is as follows: the first type of node maintains its current working channel unchanged in order to ensure that the backbone topology does not oscillate.

3. The channel selection method for distributed wireless ad hoc networks according to claim 1, characterized in that, The occupied time slot is determined by any of the following methods: The system checks whether a valid preamble of a preset format exists within a time slot to be included in the network's service usage. If the received signal strength reaches a preset occupancy threshold and no valid preamble is detected within the detection time slot, it is counted as external interference occupancy. The channel busyness index is the sum of the proportions of various occupied time slots obtained statistically over multiple consecutive superframe periods.

4. The channel selection method for distributed wireless ad hoc networks according to claim 2, characterized in that, For Category II nodes in the initial network access phase, when traversing all candidate channels, the target working channel that simultaneously meets the following conditions should be selected first: There is broadcasting from the upper-level network node, the channel busyness index is lower than the preset threshold, and the signal strength parameters meet the communication quality requirements; If no channel meets all of the above conditions, a new network element will be established on an idle channel as a second-type node.

5. The channel selection method for distributed wireless ad hoc networks according to claim 2, characterized in that, The channel switching conditions include: The first scenario is when the channel busyness index of a network-connected node exceeds the preset busy threshold for multiple consecutive superframes, and is further classified as either service-induced busyness or interference-induced busyness: if the proportion of parseable service frames is higher, it is determined to be service-induced busyness, and the first type of node does not hop frequencies, while the second type of node hops frequencies; if the proportion of unparseable high-energy signals is higher, it is determined to be interference-induced busyness, and both the first type of node and the second type of node hop frequencies. The second scenario is when the broadcast time slot cluster of an already connected node has a busy level of 0 and the upper-level network node is disconnected, which is determined to be a disconnected state. The second type of node is allowed to perform frequency hopping search of existing networks according to the preset channel sequence.

6. The channel selection method for a distributed wireless ad hoc network according to claim 1, characterized in that, The preset channel sequence is generated in one of the following ways: A fixed traversal order based on ascending or descending order of the channel center frequency within the operating frequency band; When persistent interference is detected, the channel list is divided into two groups, low-frequency band and high-frequency band, according to the interference frequency band, and channels outside the interference frequency band are selected alternately.

7. The channel selection method for a distributed wireless ad hoc network according to claim 1, characterized in that, The superframe is divided into a broadcast time slot cluster and a time synchronization time slot cluster. The switching of the radio frequency operating frequency is completed within a predetermined transition period after the broadcast time slot cluster or the time synchronization time slot cluster ends, without interrupting the service transmission of the data transmission time slot cluster.

8. The channel selection method for a distributed wireless ad hoc network according to claim 7, characterized in that, The time slot occupancy of the broadcast time slot cluster, time synchronization time slot cluster, and data interaction time slot cluster is statistically analyzed independently to obtain the busyness index of each time slot cluster dimension. The busyness index of each time slot cluster dimension is used to help determine the channel availability level.

9. An electronic device, characterized in that, It includes at least one processor and a memory connected to the processor, wherein: The memory is used to store computer programs; When the processor executes the computer program, it implements the channel selection method for a distributed wireless ad hoc network as described in any one of claims 1 to 8.

10. A computer storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the channel selection method for a distributed wireless ad hoc network as described in any one of claims 1 to 8.

Images