Hospital campus wireless network system supporting internet of things protocol

Abstract

The application relates to the technical field of wireless communication, in particular to a hospital area wireless network system supporting an Internet of Things protocol, which comprises the following: a multi-protocol dynamic scheduling engine intelligently selects a protocol stack by using a three-layer dynamic decision architecture and a lightweight reinforcement learning model; a time-sensitive hybrid topology network provides a star network and low-delay time slot allocation for emergency equipment by adopting dual-mode topology fusion; a cross-protocol security aggregation gateway realizes cross-protocol data encapsulation and field-level mapping based on a unified data header of a medical Internet of Things; and an intelligent spectrum anti-interference unit obtains spectrum sensing information results by monitoring and analyzing a hospital frequency band electromagnetic environment in real time, and feeds back the spectrum sensing information results to the multi-protocol dynamic scheduling engine and the time-sensitive hybrid topology network. The application can solve the communication problems of equipment types and complex environments by combining the multi-protocol dynamic scheduling engine with the time-sensitive hybrid topology network.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of wireless communication, in particular to a hospital campus wireless network system supporting an Internet of Things protocol. BACKGROUND

[0002] The hospital campus wireless network system supporting the Internet of Things protocol aims to optimize multi-protocol compatibility and improve network resource allocation efficiency. By combining a lightweight reinforcement learning model with a multi-protocol dynamic scheduling engine, the protocol stack selection and network topology of medical devices are controlled, the network resources of medical devices are intelligently scheduled in real time, the priority transmission of emergency medical devices is ensured, and the stability and reliability of non-emergency devices are guaranteed.

[0003] The existing hospital campus wireless network system usually causes network congestion during emergency medical data transmission. Due to the large variety of hospital devices, complex network environment and unstable channel quality, the medical devices cannot be connected in time under high load, the transmission delay is increased, and even data packet loss occurs. Therefore, a hospital campus wireless network system supporting an Internet of Things protocol is designed. SUMMARY

[0004] The application aims to provide a hospital campus wireless network system supporting an Internet of Things protocol to solve the problems of medical devices that cannot be connected in time under high load, increased transmission delay and even data packet loss due to the large variety of hospital devices, complex network environment and unstable channel quality.

[0005] A hospital campus wireless network system supporting an Internet of Things protocol comprises:

[0006] A multi-protocol dynamic scheduling engine adopts a three-layer dynamic decision-making architecture to intelligently select a protocol stack through a lightweight reinforcement learning model.

[0007] A time-sensitive hybrid topology network adopts a dual-mode topology fusion to establish a star network for emergency medical devices and allocate low-delay time slots, and to construct a self-organizing Mesh network for non-emergency devices and dynamically optimize multi-hop transmission paths.

[0008] A cross-protocol security aggregation gateway integrates a medical Internet of Things unified data header protocol to realize field-level mapping and packaging of multi-protocol data, and performs dynamic token authentication based on a national cryptographic algorithm SM9 and a zero-trust architecture.

[0009] An intelligent spectrum anti-interference unit monitors hospital frequency bands in real time through software-defined radio, generates spectrum sensing results and feeds them back to the multi-protocol dynamic scheduling engine and the time-sensitive hybrid topology network to adjust network configuration.

[0010] As a further improvement of the technical solution, the multi-protocol dynamic scheduling engine implements multi-protocol dynamic scheduling based on hierarchical decision-making, including that the multi-protocol dynamic scheduling engine comprises:

[0011] a device layer, which identifies device types and assigns initial protocol stacks through embedded protocol tags;

[0012] a network layer, which dynamically switches protocol stacks based on real-time channel quality evaluation, wherein the channel quality evaluation adopts a lightweight reinforcement learning model;

[0013] a service layer, which binds protocol stacks with quality of service (QoS) according to medical data priorities.

[0014] As a further improvement of the technical solution, the protocol stack dynamic switching method of the network layer comprises:

[0015] constructing a network quality evaluation function to calculate the channel quality of the initial protocol stack based on signal strength, packet loss rate and delay;

[0016] when the channel quality is lower than a threshold value, constructing a network state space and an action space, and selecting an optimal protocol stack through a reinforcement learning model;

[0017] adopting a reward function to combine signal strength, packet loss rate, delay and protocol stack characteristics, updating Q values of the protocol stack based on the network state space, and iteratively optimizing protocol stack decisions.

[0018] As a further improvement of the technical solution, the service layer is configured to:

[0019] assign a low-latency high-reliability protocol stack to emergency device data streams;

[0020] assign a low-bandwidth low-power protocol stack to regular monitoring device data streams.

[0021] As a further improvement of the technical solution, the time-sensitive hybrid topology network comprises:

[0022] a star time slot preemption module, which calculates preemptive time slot assignments based on device dynamic weights;

[0023] a Mesh path optimization module, which calculates optimal multi-hop transmission paths through link quality indices and energy remaining.

[0024] As a further improvement of the technical solution, the star time slot preemption module performs:

[0025] calculates dynamic weights according to device priority labels, signal-to-noise ratios and signal strengths;

[0026] The time slot length is based on the weight proportion distribution theory, and time slot preemption is performed on high-priority devices.

[0027] As a further improvement of the technical solution, the Mesh path optimization module performs:

[0028] The link quality index between the computing devices is calculated, and the signal strength, error rate and hop count parameters are fused;

[0029] The optimal multi-hop transmission path is selected based on the principle of minimizing the comprehensive cost.

[0030] As a further improvement of the technical solution, the cross-protocol security aggregation gateway comprises:

[0031] The protocol data mapping module realizes multi-protocol data field mapping and time window aggregation through a unified data header structure;

[0032] The dynamic trust chain authentication module generates a short-term access token based on an elliptic curve password and performs context-aware verification.

[0033] As a further improvement of the technical solution, the unified data header structure contains device identification, data type coding, time stamp and verification field, and the MQTT Topic and Zigbee Cluster ID are mapped to standard coding through protocol analysis rules.

[0034] As a further improvement of the technical solution, the intelligent spectrum anti-interference unit comprises:

[0035] The spectrum anti-interference module is used for real-time scanning of the electromagnetic environment and generating a frequency band interference map;

[0036] The sensing information feedback module is used for triggering protocol stack switching and network topology reconstruction according to the interference map.

[0037] Compared with the prior art, the beneficial effects of the present application are:

[0038] 1、The present application is based on a multi-protocol cooperative scheduling algorithm and a network resource optimization strategy, which can intelligently select a suitable protocol stack according to the real-time state of the device, data transmission requirements and network load, and dynamically adjust the allocation of network resources, thereby improving the connection efficiency of the device, reducing the delay, guaranteeing the transmission channel of high-priority medical devices, and optimizing the bandwidth usage of low-priority devices.

[0039] 2、The present application introduces a network topology dynamic adjustment mechanism based on deep reinforcement learning, realizes real-time monitoring of device distribution, network quality and channel state, automatically optimizes the network topology structure, ensures that the medical devices can be quickly and stably connected to the network under different areas and different loads, avoids network congestion, and ensures efficient and stable data transmission. BRIEF DESCRIPTION OF DRAWINGS

[0040] Figure 1 is a whole flow chart of the present application;

[0041] The meanings of various labels in the figure are as follows:

[0042] 1, multi-protocol dynamic scheduling engine; 11, device layer; 12, network layer; 13, service layer; 2, time-sensitive hybrid topology network; 21, star time slot preemption module; 22, Mesh path optimization module; 3, cross-protocol security aggregation gateway; 31, protocol data mapping module; 32, dynamic trust chain authentication module; 4, intelligent spectrum anti-jamming unit; 41, spectrum anti-jamming module; 42, perception information feedback module. DETAILED DESCRIPTION

[0043] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0044] Please refer to Figure 1 As shown in the figure, a hospital campus wireless network system supporting Internet of Things protocols is provided, comprising:

[0045] A multi-protocol dynamic scheduling engine 1, which intelligently selects a protocol stack by a three-layer dynamic decision architecture using a lightweight reinforcement learning model;

[0046] The multi-protocol dynamic scheduling engine 1 realizes multi-protocol dynamic scheduling based on hierarchical decision, including a device layer 11, a network layer 12 and a service layer 13;

[0047] The device layer 11 automatically identifies device types through embedded protocol labels and allocates initial protocol stacks to devices according to device types; a life monitor is automatically marked as an MQTT+Urgent protocol to ensure low latency and reliable transmission; an environmental sensor is marked as a CoAP+LowPower protocol to adapt to low-power and low-bandwidth requirements;

[0048] The network layer 12 uses a lightweight reinforcement learning model to evaluate channel quality in real time and dynamically switch device protocol stacks;

[0049] The reinforcement learning model is a machine learning technology applied in multiple fields, including network optimization, autonomous driving, robot control, etc.; the lightweight reinforcement learning model, as a reinforcement learning model optimized for resource-limited devices, is a simplified and optimized form of reinforcement learning model, which is suitable for edge computing and Internet of Things scenarios of the hospital campus wireless network.

[0050] The service layer 13 assigns a transmission strategy according to the medical data priority, and strongly binds the device protocol stack with QoS.

[0051] In the embodiment, the network layer 12 evaluates the channel quality in real time by using a lightweight reinforcement learning model, and dynamically switches the device protocol stack. The specific method steps are as follows:

[0052] S1.2.1, based on signal strength, packet loss rate and delay, construct network quality evaluation function, calculate initial protocol stack channel quality:

[0053] Normalize each index:

[0054] Map the signal strength to the interval [0, 1]:

[0055] Wherein, The worst signal, The best signal.

[0056] Packet loss rate is inversely mapped:

[0057]

[0058] Delay is inversely mapped:

[0059]

[0060] Wherein, The maximum acceptable delay.

[0061] Network quality evaluation function:

[0062] ;

[0063] Wherein, The initial protocol stack channel quality; The signal strength weight coefficient; The signal strength; The packet loss rate weight coefficient; The packet loss rate; The delay weight coefficient; The delay, , , The normalized values of signal strength, packet loss rate and delay respectively. In the embodiment, .

[0064] In the embodiment, before dynamic protocol switching, it is first necessary to evaluate the channel quality of the network in real time to determine whether the protocol stack needs to be switched; the evaluation of network quality includes signal strength, packet loss rate and delay;

[0065] S1.2.2, if the channel quality of the initial protocol stack is lower than the device requirement standard, dynamically switch the device protocol stack to perform S1.2.3; if the channel quality of the initial protocol stack meets the device requirement standard, perform real-time continuous monitoring;

[0066] S1.2.3, construct the network state space at time t based on signal strength, packet loss rate and delay:

[0067] ;

[0068] wherein, is the current time; is the network state space at time t; is the signal strength at time t; is the packet loss rate at time t; is the delay at time t;

[0069] S1.2.4, construct the network action space at time t based on the network state space at time t and the initial protocol stack:

[0070] ;

[0071] wherein, is the selected protocol stack at time t; is the Q value of the protocol stack under the network state space at time t, indicating the expected benefit of the protocol stack in the current network environment, is the set of protocol stacks available to the current device.

[0072] S1.2.5, construct the reward function:

[0073] ;

[0074] wherein, is the reward function at time t; is the natural influence of the protocol stack on the reward, which is automatically generated through its throughput, power consumption and other characteristics; specifically as follows:

[0075]

[0076] wherein, is the measured throughput (unit: Mbps) of the protocol stack , is the transmission success rate of the protocol stack , ranging from [0, 1], is the power consumption (unit: mW) of the protocol stack . , , max (all protocol stacks in the system) (for normalization). , throughput weight, , reliability weight, , power consumption weight.

[0077] S1.2.6, based on the network state space at time t, the selected protocol stack, and the reward function, the lightweight reinforcement learning model updates the Q value of the initial protocol stack under the network state space at time t using the Q-learning strategy:

[0078]

[0079] wherein, is the learning rate; is the discount factor; is the selected protocol stack at time t+1; is the network state space at time t+1; is the maximum Q value estimate of all protocol stacks under the network state space at time t+1; is the maximum Q value brought by executing the protocol stack at time t; , 0.9≤ ≤0.99.

[0080] S1.2.7, based on the Q value of the protocol stack under the network state space at time t, the lightweight reinforcement learning model selects the protocol stack with the maximum Q value according to the network state space at time t .

[0081] In this embodiment, the service layer 13 assigns a transmission strategy according to the medical data priority, and strongly binds the device protocol stack with QoS. The specific method steps are as follows:

[0082] S1.3.1, according to the medical data priority, the low-latency high-reliability protocol stack is assigned to high-priority medical data;

[0083] S1.3.2, according to the medical data priority, the low-bandwidth low-power protocol stack is assigned to low-priority medical data.

[0084] High-priority medical data includes emergency instructions, real-time vital sign data, and emergency medical reports, etc.; low-priority medical data includes log uploading, daily health data, and non-emergency examination reports, etc.

[0085] It also includes a time-sensitive hybrid topology network 2, which adopts a dual-mode topology fusion to provide a star network and a low-delay time slot allocation for emergency devices, and a multi-hop transmission through a Mesh relay network for non-emergency devices and dynamic signal interference avoidance.

[0086] The time-sensitive hybrid topology network 2 comprises a star time slot preemption module 21 and a Mesh path optimization module 22;

[0087] The star time slot preemption module 21 is configured to construct a low-latency star network for emergency equipment.

[0088] The Mesh path optimization module 22 constructs a self-organizing Mesh network for non-emergency equipment, dynamically optimizes multi-hop transmission paths, and avoids signal interference.

[0089] In this embodiment, the star time slot preemption module 21 is configured to construct a low-latency star network for emergency equipment, and the specific method steps are as follows:

[0090] S2.1.1, device priority dynamic registration, the device carries a priority label to send a registration request to the central controller, and the central controller calculates the dynamic weight of the device:

[0091] ;

[0092] Wherein, is the device index in the dynamic weight; is the dynamic weight of the kth device; is the priority label weight; is the priority label of the kth device, is the last priority label; is the signal-to-noise ratio weight; is the signal-to-noise ratio of the kth device; is the maximum signal-to-noise ratio; is the signal strength weight; is the signal strength of the kth device; is the signal strength threshold.

[0093] In this embodiment, the priority label is divided into five levels, i.e., first to fifth levels; represents the priority label of the kth device as the first level; represents the priority label of the kth device as the second level, and so on; the signal strength threshold is -70 dBm.

[0094] S2.1.2, based on the dynamic weight of the kth device and the time slot period, the theoretical time slot length of the kth device is calculated:

[0095] ;

[0096] Wherein, is the theoretical time slot length of the kth device; is the time slot period; is the first device index in the time slot length; is the total number of registered devices; is the dynamic weight of the th device;

[0097] If the priority of the newly registered device is lower than the priority of the th device, the time slot is forcibly allocated, realizing the pre-emptive TDMA time slot allocation function:

[0098] ;

[0099] wherein, is the actual time slot length of the th device; is the theoretical time slot length of the th device; is the set of devices with priority labels lower than the fourth level; is the second device index in the time slot length; is the theoretical time slot length of the th device.

[0100] In the present embodiment, the Mesh path optimization module 22 constructs a self-organizing Mesh network for non-urgent devices, dynamically optimizes multi-hop transmission paths and avoids signal interference, and the specific method steps are as follows:

[0101] S2.2.1, calculating ;

[0102] wherein, is the first device index in the transmission path; is the second device index in the transmission path; is the link quality index between device m and device n, ; is the signal strength between device m and device n; is the maximum signal strength, is the minimum signal strength; is the bit error rate attenuation coefficient, 2 ≤ ≤ 5; is the bit error rate between device m and device n; is the hop number between device m and device n; is the hop number penalty factor, 1 ≤ ≤ 2.

[0103] S2.2.2, based on the link quality index between devices, the optimal multi-hop transmission path for non-urgent devices is calculated, realizing the transmission path optimization function of non-urgent devices:

[0104]

[0105] wherein, is a path index; is the a-th path; is a path comprehensive cost; is an energy weight coefficient; the energy weight coefficient represents the proportion of device energy in the path; is a device energy remaining, is full energy;

[0106] ;

[0107] wherein, is an optimal multi-hop transmission path; is a set of all paths; is a path index; is the a-th path.

[0108] Further comprising a cross-protocol security aggregation gateway 3, which realizes cross-protocol data encapsulation and field-level mapping based on a medical Internet of Things unified data header, and integrates a national secret SM9 algorithm and a zero-trust architecture, dynamically allocates an access token while performing real-time verification;

[0109] The cross-protocol security aggregation gateway 3 comprises a protocol data mapping module 31 and a dynamic trust chain authentication module 32;

[0110] In the embodiment, the protocol data mapping module 31 realizes multi-protocol data field-level mapping and format normalization through a medical Internet of Things unified data header, and the specific method steps are as follows:

[0111] S3.1.1, define a unified data header structure:

[0112] ;

[0113] wherein, is a unified data header structure; is a device unique identifier; is a data type code; is a timestamp; is a cyclic redundancy check code;

[0114] S3.1.2, regular analysis is performed on the Topic field of the MQTT protocol, key information is extracted and encoded to DataClass, and the Cluster ID of the Zigbee protocol is table mapping to generate a standard DataClass code;

[0115] ​S3.1.3, aggregate data based on time window:

[0116]

[0117] wherein, is the aggregated data set; is the time window index; is the total duration of data collection; is the time window size; is the data index; is the th data; is the timestamp of the th data; is the time interval of the th time window.

[0118] In this embodiment, the dynamic trust chain authentication module 32 is based on the SM9 algorithm and the zero trust architecture, which is used for device dynamic token allocation and context-aware verification. The specific method is as follows:

[0119] S3.2.1, based on device identity, master private key and generation element, using hash to elliptic curve group function to calculate device public key, generate short-term access token;

[0120] S3.2.2, verify the validity of short-term access token, and sense the matching of verification device and user terminal.

[0121] It also includes an intelligent spectrum anti-interference unit 4, which obtains spectrum sensing information results by real-time monitoring and analyzing the electromagnetic environment of the hospital frequency band through software defined radio technology, and feeds back the spectrum sensing information results to the multi-protocol dynamic scheduling engine 1 and the time-sensitive hybrid topology network 2.

[0122] The intelligent spectrum anti-interference unit 4 includes an electromagnetic spectrum anti-interference module 41 and a sensing information feedback module 42, which are specifically as follows:

[0123] In this embodiment, the spectrum anti-interference module 41 is used to collect electromagnetic spectrum data in real time, analyze and evaluate the electromagnetic spectrum data to obtain spectrum sensing information results, adjust the frequency band resources and update the working frequency band configuration of the system.

[0124] Among them, the sensing information data results include signal strength, interference evaluation and frequency band selection, etc.

[0125] In this embodiment, the sensing information feedback module 42 is used to feed back the spectrum sensing information results to the multi-protocol dynamic scheduling engine 1 and the time-sensitive hybrid topology network 2, and reconfigure the network by dynamically switching the protocol stack.

[0126] The above shows and describes the basic principles, main features and advantages of the present application. Those skilled in the art should understand that the present application is not limited to the above-mentioned embodiments, and the above-mentioned embodiments and descriptions in the specification are only preferred examples of the present application and are not intended to limit the present application. Without departing from the spirit and scope of the present application, various changes and improvements can be made to the present application, and these changes and improvements all fall within the scope of the claimed present application.

Claims

1. A hospital campus wireless network system supporting Internet of Things (IoT) protocols, characterized in that, include: A multi-protocol dynamic scheduling engine (1) is provided, which adopts a three-layer dynamic decision architecture and uses a lightweight reinforcement learning model to intelligently select the protocol stack. The three-layer dynamic decision-making architecture includes: The device layer (11) identifies the device type and allocates the initial protocol stack through embedded protocol tags; The network layer (12) dynamically switches the protocol stack based on real-time channel quality assessment, where the channel quality assessment adopts the network quality assessment function; The business layer (13) binds the protocol stack to the Quality of Service (QoS) based on the priority of medical data; The time-sensitive hybrid topology network (2) adopts dual-mode topology fusion to establish a star network for emergency medical equipment and allocate low-latency time slots, and builds a self-organizing mesh network for non-emergency equipment and dynamically optimizes multi-hop transmission paths; Cross-protocol security aggregation gateway (3), the cross-protocol security aggregation gateway (3) integrates the unified data header protocol of medical IoT, realizes field-level mapping and encapsulation of multi-protocol data, and performs dynamic token authentication based on the national cryptographic SM9 algorithm and zero trust architecture; The intelligent spectrum anti-interference unit (4) monitors the hospital frequency band in real time through software-defined radio, generates spectrum sensing results and feeds them back to the multi-protocol dynamic scheduling engine (1) and the time-sensitive hybrid topology network (2) to adjust the network configuration.

2. The hospital campus wireless network system supporting the Internet of Things protocol according to claim 1, characterized in that: The protocol stack dynamic switching method of the network layer (12) includes: Construct a network quality assessment function to calculate the channel quality of the initial protocol stack based on signal strength, packet loss rate, and delay; When the channel quality is below the threshold, a network state space and action space are constructed, and the optimal protocol stack is selected through a reinforcement learning model. By combining the reward function with signal strength, packet loss rate, latency and protocol stack characteristics, and based on the Q value of the protocol stack in the network state space, the Q value is updated and the protocol stack decision is iteratively optimized.

3. The hospital campus wireless network system supporting the Internet of Things protocol according to claim 2, characterized in that: The business layer (13) is configured as follows: Distribute low-latency, high-reliability protocol stacks to the data stream of emergency medical equipment; Distribute low-bandwidth, low-power protocol stacks to the data streams of conventional monitoring devices.

4. The hospital campus wireless network system supporting the Internet of Things protocol according to claim 3, characterized in that: The time-sensitive hybrid topology network (2) includes: The star-shaped time slot preemption module (21) allocates time slots based on the dynamic weight calculation of the equipment. The Mesh path optimization module (22) calculates the optimal multi-hop transmission path using the link quality index and remaining energy.

5. The hospital campus wireless network system supporting the Internet of Things protocol according to claim 4, characterized in that: The star-shaped time slot preemption module (21) executes: Dynamic weights are calculated based on device priority labels, signal-to-noise ratio, and signal strength. The time slot length is allocated based on the weight ratio theory, and time slot preemption is performed on high-priority devices.

6. The hospital campus wireless network system supporting the Internet of Things protocol according to claim 4, characterized in that: The Mesh path optimization module (22) performs the following: Calculate the link quality index between devices, and integrate signal strength, bit error rate and hop count parameters; The optimal multi-hop transmission path is selected based on the principle of minimizing the overall path cost.

7. The hospital campus wireless network system supporting the Internet of Things protocol according to claim 1, characterized in that: The cross-protocol security aggregation gateway (3) includes: The protocol data mapping module (31) realizes multi-protocol data field mapping and time window aggregation through a unified data header structure; The dynamic trust chain authentication module (32) generates short-term access tokens based on elliptic curve cryptography and performs context-aware verification.

8. The hospital campus wireless network system supporting the Internet of Things protocol according to claim 7, characterized in that: The unified data header structure includes device identifier, data type encoding, timestamp and verification field, and maps MQTT Topic and Zigbee Cluster ID to standard encoding through protocol parsing rules.

9. The hospital campus wireless network system supporting the Internet of Things protocol according to claim 8, characterized in that: The intelligent spectrum anti-interference unit (4) includes: The spectrum anti-interference module (41) is used to scan the electromagnetic environment in real time and generate frequency band interference maps; The perception information feedback module (42) is used to trigger protocol stack switching and network topology reconstruction based on the interference map.

Images