High-elasticity intelligent voice calling system

Abstract

The invention belongs to the field of voice calling, and relates to a high-elasticity intelligent voice calling system, which adopts Keepalived software to realize high availability through a VRRP (Virtual Router Redundancy Protocol), and comprises a PBX (Private Branch Exchange) calling system; a JCWI service cloud module; hSSVP-BE and HSSVP-FE are adopted in the method. According to the invention, a plurality of voice calling systems of each office are combined and interconnected by using a cloud computing technology, so that high scalability is achieved; meanwhile, priority ranking is carried out on user queues based on the sensitive levels of the calling users, and the service inclusiveness is enhanced.

Description

Technical Field

[0001] The invention belongs to the field of voice calling and relates to a highly elastic intelligent voice calling system. Background Art

[0002] An important part of the construction of the railway information system collaborative work platform is the construction of a voice call platform, which is used to broadcast voice notifications, urgently communicate important information, coordinate work items, unify resource dispatch, and command personnel, to ensure that in actions involving multiple people, instructions are issued quickly, personnel respond quickly, and the team moves in unison. This type of system is particularly suitable for emergency response, dispatching and command and other sudden scenarios.

[0003] The construction of voice call platform systems will encounter the problem of difficulty in matching system planning capacity with actual usage needs. The handling of emergencies is not a daily job. If the planned capacity basically matches the emergency demand, a large amount of capacity will be idle on a daily basis, resulting in waste; if it only meets the daily usage needs, when an emergency occurs, it will be overwhelmed and difficult to cope with. For example, a voice call pool is planned for 30 concurrent voice calls according to daily use. It seems to be easy to use for small-scale daily applications. Once it encounters large-scale applications, when calling 2,000 or more people at the same time, each person takes an average of 3 minutes to connect, and it takes at least 200 minutes, or more than 3 hours, to complete the call, and the response effect of personnel is greatly reduced. If it is changed to planning 1,000 channels, the capacity will be much stronger when encountering large-scale calls, but it requires a lot of investment and operating costs, but it is often not fully used on a daily basis, resulting in waste. Summary of the invention

[0004] In view of the above problems, the present invention provides a highly flexible intelligent voice call system which can unite multiple voice call systems of various offices and interconnect them using cloud computing technology to achieve high scalability.

[0005] In order to achieve the above object, the technical solution adopted by the present invention is:

[0006] A highly flexible intelligent voice call system that uses Keepalived software to achieve high availability through the VRRP protocol, including:

[0007] PBX calling system;

[0008] JCWI Business Cloud Module;

[0009] HSVCP-BE and HSVCP-FE are the front-end and back-end of the WEB management system.

[0010] Preferably, the PBX call system is mainly responsible for rapid text-to-speech conversion, service call notification, information collection, supports intelligent routing, call service intelligent queues, supports response level and number management, and can achieve multi-machine load balancing.

[0011] Preferably, the JCWI business cloud module is mainly responsible for the communication between the business system and the voice call platform, and the shared services between clouds.

[0012] Preferably, VRRP is Virtual Router Redundancy Protocol.

[0013] Preferably, the PBX call system adopts a dual-path architecture of an application server and a call server.

[0014] Preferably, the PBX calling system includes a voice calling platform and a background management system.

[0015] Preferably, the voice call platform includes a business interface, JUSTTTS voice conversion, intelligent routing, JUSTIVR interactive voice response, underlying hardware interface and cloud function.

[0016] Preferably, the background management system includes system management, application management, parameter configuration, statistical analysis, billing management and system monitoring.

[0017] Preferably, the JCWI business cloud module adopts cloud computing architecture technology, which is divided into two levels: first, the local system is built using a K8S application server cluster, and various network computing resources are used for load balancing and automatic fault transfer, which is equivalent to building a local cloud; second, the local clouds of each bureau are interconnected using Huawei Kurator for cluster management, making it easy to expand the cluster group. After the local cloud of a new bureau joins the cluster group, it immediately shares its own resources and simultaneously obtains the shared resources of other bureaus in the group.

[0018] Compared with the prior art, the advantages and positive effects of the present invention are:

[0019] 1. The present invention studies highly elastic voice call technology and transforms the existing voice platform into a highly available shared service for important business calls. When the voice call pool has limited resources, the present invention automatically and intelligently rents similar call resources from the associated call pool and combines them together to cope with sudden surges in access requests, thereby achieving high scalability and reflecting the core concept of "high elasticity". At the same time, the system functions are further packaged into shared services for numerous information systems to call, reflecting high inclusiveness and flexibility, while also being safe and controllable. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0021] Figure 1 This is the overall architecture diagram of the highly flexible intelligent voice call system;

[0022] Figure 2 This is a high availability architecture diagram of the database for a highly elastic intelligent voice call system;

[0023] Figure 3 This is a load balancing architecture diagram for each component of the highly elastic intelligent voice call system;

[0024] Figure 4 This is a system functional block diagram of a highly flexible intelligent voice call system;

[0025] Figure 5 Extend the interconnection graph for cloudification among multiple systems;

[0026] Figure 6 This is a simple diagram sorted by profession;

[0027] Figure 7 Schematic diagram of the queue for grouping sensitivity levels;

[0028] Figure 8 This is a monitoring diagram for call connection status. DETAILED DESCRIPTION

[0029] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described below in conjunction with the accompanying drawings and embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0030] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways than those described herein. Therefore, the present invention is not limited to the specific embodiments of the following disclosure.

[0031] Embodiment 1, as Figure 1-Figure 5 As shown, the present invention provides a highly elastic intelligent voice call system. The entire system is a multi-cloud interconnected structure and can be dynamically expanded according to demand. The main components are described as follows:

[0032] (1) PBX call system: It is mainly responsible for tasks such as rapid text-to-speech conversion, service call notification, and information collection. It supports intelligent routing, call service intelligent queues, response level and number management, and can achieve multi-machine load balancing.

[0033] (2) JCWI business cloud module: mainly responsible for the communication between the business system and the voice call platform, as well as the shared services between clouds.

[0034] (3)HSVCP-BE, HSVCP-FE: front-end and back-end of the WEB management system.

[0035] like Figure 2 As shown, in this system, Keepalive software is used to achieve high availability through VRRP (Virtua l Router Redundancy Protocol) protocol, shortening the database downtime caused by daily maintenance operations (planned) and sudden system crashes (unplanned), so as to improve the overall system availability.

[0036] like Figure 3 As shown in the figure, both the application server and the call server adopt a dual-path architecture, and the high-performance reverse proxy software Ngi nx plays the core role of the load balancing architecture at the bottom layer.

[0037] like Figure 4 As shown, this system is divided into two parts according to its functional design: voice call platform and background management system.

[0038] 1. Voice Call Platform

[0039] Receive call requests from upper-layer applications, complete outbound calls through the lines of communication operators, notify upper-layer applications of the call results, and record relevant data of the entire process, mainly including:

[0040] 1. Business interface: The voice call platform provides a variety of interfaces for internal business systems. Business users submit outbound tasks through the http interface, and can specify different business numbers, divide them into levels, and jump the queue. Other functions provided include:

[0041] 1) Cancel notification: If the waiting call queue is too long, you do not want to wait any longer or there is an error in operation, you can choose to cancel the submitted outbound task. You can cancel by batch or single number.

[0042] 2) Get real-time information about the queue: including the time required to complete the call task, the number of calls dialed in this batch, the number of calls not dialed in this batch, and the number of calls still needed to complete the task.

[0043] 2. JUSTTTS voice conversion: convert notification text content into voice, convert text notification content into voice for later download or review.

[0044] 3. Intelligent routing:

[0045] 1) To achieve simultaneous submission of multiple queue tasks, when urgent large-scale multi-region notification tasks are required and the notification scale is far greater than the system capacity, in order to allow each region, each major relevant department or personnel to be notified in time, the business user needs to initiate multiple queue notification tasks at one time, so that the phone number of each queue can receive polling notifications. Multiple queues can be used to jump the queue. When a large number of notification tasks are waiting in the queue for execution, if there are existing urgent tasks that need to be submitted and hope to be notified as soon as possible, a new queue can be created to jump the queue.

[0046] 2) Outbound call line resources are dynamically allocated by group. The number of available relays on the platform is dynamically and evenly distributed according to the group of outbound tasks, so that call tasks of different groups can be handled evenly.

[0047] 3) Relays are allocated based on level priority. The platform's available lines allocate line resources to high-level tasks based on the priority of the outbound task level. Low-priority outbound tasks are suspended, and notification messages are pushed to the low-priority business initiators. A number of outbound call resources are reserved for high-priority tasks, and this number of line resources will not be used by low-level businesses, so that high-priority tasks can be executed immediately after submission.

[0048] 4. JUST IVR interactive voice response: Automatically realize voice interaction between users and the system, and obtain user feedback on notification content. Users can actively call in and automatically obtain relevant information.

[0049] 5. Bottom-level hardware interface: Integrate and match voice communication equipment from multiple communication manufacturers, including Hangzhou Synway and Huawei switches.

[0050] 6. Cloud function: Combine cloud technology to realize cloud sharing of call resources.

[0051] In order to solve the problem of enthusiasm and fairness in providing resource sharing services to each other, a billing module is planned and designed to support the settlement of communication fees and service fees between each other. The interconnected parties can earn revenue by providing services to the outside world to make up for their own construction investment and operation and maintenance costs. The multimedia call platform collects call information in real time for statistical billing, including: phone number, call duration, etc.

[0052] 2. Backend Management System

[0053] The backend management system is the operation portal of the entire system, providing system backend management functions to realize main functions such as system management, application management, parameter configuration, statistical analysis, billing management, and system monitoring.

[0054] 1. System management: The functions implemented include parameter configuration, providing a system parameter configuration function interface; account management, business user account and other information management, including add, delete, modify, query and other operations.

[0055] 2. Application management: Provides system access application maintenance functions. Including:

[0056] 1) Application authorization: You can manage the application permissions connected to the system. Click a specific application in the application list to enter the application authorization interface.

[0057] 2) Application priority management: manage the priority of applications accessing the system.

[0058] 3) Application resource allocation: allocate available line resources to applications.

[0059] 3. Statistical management: Provides outbound call statistics of the application, provides application call detail query function, etc.

[0060] 4. Billing management: You can define the billing rules of the system and provide application fee details query and settlement functions.

[0061] 5. System monitoring: Real-time display of various software and hardware status in the system, including relay status, call queue, subsystem program status, etc.

[0062] 3. Key technologies

[0063] It is reflected in two aspects: 1) The cloud-based extended interconnection technology of the call system is used to ensure high elasticity; 2) The technology is used to improve the user experience and enhance the response effect in order to reflect intelligence.

[0064] 3.1 Call system cloud expansion technology

[0065] Traditional systems are interconnected in pairs, and the entire platform has many connection channels. Development and operation and maintenance are extremely complex, and the expansion of call capacity and capabilities is not easy. In order to achieve high elasticity and interconnection convenience, cloud computing architecture and technology are used to connect the local call systems of each bureau. Cloud computing technology ensures high elasticity and scalability of platform capabilities, thereby reflecting high availability. It is divided into two levels: First, the local system is built using K8S application server clusters, and various network computing resources are used for load balancing and automatic fault transfer, which is equivalent to building a local cloud; second, the interconnection between the local clouds of each bureau uses Huawei Kurator for cluster management, which makes it easy to expand the cluster group. After the local cloud of a new bureau joins the cluster group, it immediately shares its own resources and simultaneously obtains the shared resources of other bureaus in the group. For details, Figure 5 shown.

[0066] 3.2 Technologies to improve response effects

[0067] When resources are still tight after cloud expansion, we should focus on taking care of users' psychology and experience, adopt appropriate technologies, use resources where they are needed most, ensure that units and organizations are given priority, and process some users who are less sensitive to emergencies post-processing, and induce users to adapt themselves and cooperate in supportive behaviors, thereby effectively alleviating users' anxiety and discomfort while waiting.

[0068] 3.2.1 User Grouping

[0069] After a call channel receives a call task and starts operating, it cannot be forcibly interrupted and must wait until the call is completed before it can be used again. Because of this exclusive exclusivity, when using call resources, different priorities should be marked for tasks of different urgency. As long as it is still in the pending queue and has not been sent to the call channel, a high-priority task[5] can be queued in front of a low-priority task to obtain the idle call resources first and get priority service. Generally speaking, the more serious the potential damage caused by the event and the larger the impact range, the higher its priority is arranged. By default, the priority of the group-level first-level response event is set as the highest level, followed by the group-level second-level and third-level responses, and then the station-level events.

[0070] Considering the sporadic nature of emergencies, it is generally believed that if a large-scale emergency occurs in a certain region, it will not occur in other places at the same time. The probability of multiple large-scale emergencies occurring at the same time is very low. In this case, based on the jurisdiction of the railway bureau, we can invest certain resources in different regions to build local voice call systems for each railway bureau, undertake some of their daily business, and occupy some resources; at the same time, let the voice call systems of each bureau be interconnected [6], share the idle and redundant resources, and adjust them for use.

[0071] The above ideas alone are not enough. For example, in a large-scale emergency call initiated by the Guangzhou Bureau, 20,000 people were involved in the simultaneous calls. The five bureaus in Guangzhou, Chengdu, Lanzhou, Shenyang, and Jinan each had 150 lines available for sharing, and the 50 lines reserved by the bureau itself, totaling 800 lines. These 800 lines called 20,000 people, and each person took an average of 3 minutes to get through. It took 75 minutes to complete the call. The first person to know the situation was 75 minutes faster than the last person to know. With so many people, it was impossible to achieve unified command and unify the pace. Therefore, we have to find a way to solve the problem of never having enough resources. We can tap the potential from the characteristics of the call service itself and reduce the sensitivity caused by the shortage of call resources.

[0072] In an emergency call, suppose the users notified involve six professional grassroots stations and sections: vehicle maintenance, engineering maintenance, electrical maintenance, power supply, locomotive maintenance, and vehicle maintenance. If we simply sort them by common professions, stations and personnel, such as Figure 6 shown.

[0073] The final result is that the personnel in the vehicle affairs department are the first to receive the notification, the vehicles are the most delayed, the notifications received by different professions are not synchronized, and work cannot be carried out in a coordinated manner.

[0074] If the six groups are separated into vehicle maintenance, engineering maintenance, electrical maintenance, power supply, locomotive maintenance, and vehicle maintenance, and each professional group has different stations and sections, the callers are divided into three levels according to the time sensitivity of receiving the notification: the dispatch duty room (station and section emergency command center) is level A, the station and section leaders are level B, and the workshop and work area management personnel near the accident site are level C. Then, according to the sensitivity level, profession, and station section, a unified queue is constructed by unified sorting, such as Figure 7 shown.

[0075] By initiating calls according to the above-mentioned unified queue, we can ensure that when resources are limited, there will always be some people in each profession and section who receive the notification first, some people who receive the notification later, and some people who receive the notification last. In other words, when resources are tight, everyone will share the tension evenly, so as to avoid serious imbalance and too large a gap between the first and the last, which will affect the subsequent unified dispatch and command.

[0076] According to the relevance of emergency response work and profession, a priority code matrix of the organization to be called is constructed according to the professional relevance and sensitivity level, as shown in the following table:

[0077] Table 1 - User group priority code table

[0078]

[0079] When the software is designed and implemented, the priority code in the above table is placed in the first position of the sorting rules for the personnel queues sent to the call platform.

[0080] The use of user grouping and priority care for group organizations is essentially a result of a shortage of call resources. When resource shortages are perceived, users will be prompted to adapt and cooperate to make up for the shortage. The duty room staff who first receive notifications at the station level have their own fixed and mature notification channels for station-level emergency events. They will follow traditional routines and expand the scope of notifications.

[0081] For global emergency drills and large-scale emergency responses, a regional variable is introduced in the call grouping strategy to ensure that in the event of resource constraints, key priority call recipients in each region can receive notifications as soon as possible.

[0082] 3.2.2 Priority Queue Jumping

[0083] High-priority call tasks are allowed to be placed in front of the low-priority call task queue. At the same time, notifications are actively sent to low-priority users, and real-time prompts are displayed on the user interface to relieve users' anxiety and depression, thereby improving user experience.

[0084] Four application scenarios and effects

[0085] This system has been successfully applied in the Guangzhou Bureau's intelligent emergency command platform. Through intelligent, efficient and unified voice calls, it supports the primary link of emergency command work - quickly notifying and mobilizing personnel at the beginning of an emergency event - and achieves the command effect of "strategizing and responding to one call". It is explained in the following two aspects.

[0086] 4.1 Local system efficient voice calls

[0087] 4.1.1 Accurately find people

[0088] The text content of the emergency notification is automatically converted into a voice message within 1-3 seconds through iFlytek's text-to-speech and Jiashida TTS system, and accurately sent to various professional on-duty personnel and supervisors near or closely related to the incident. The notification scope is strictly controlled according to the emergency response registration, and users are notified in a strong reminder manner. In a tense and hurried situation, users do not need to bother reading cumbersome text content, and can easily obtain key information through hearing. If the user is not connected the first time, follow-up calls will be made later, with a total of multiple calls at intervals of up to five minutes (configurable by parameters). This is the rhythm of waking up and summoning necessary personnel during emergency response.

[0089] 4.1.2 Information traceability

[0090] In addition to receiving a voice call notification, users will also receive a corresponding text message and can interactively retrieve the most recent voice broadcast by calling back the call number. Supervisors who are assigned to rush to the scene can also open the mobile app to see the location of the emergency event and navigate to it with one click. When users are in a hurry, they will not miss any details because they don't remember them, which further ensures the accuracy of emergency information notifications.

[0091] 4.1.3 Differentiating sensitivity

[0092] Setting up multiple caller numbers for calls can distinguish the importance and urgency of calls and the scope of work. Dividing them into group level, dispatching office level, and station level is conducive to users choosing the processing branch according to the response level, investing corresponding support and cooperation efforts, strengthening business coordination, and ensuring that responsibilities are in place.

[0093] 4.1.4 Anti-cell phone interception

[0094] When the mobile App is launched for the first time after installation, a new contact will be created with the user’s consent, and the caller number called by the user will be added to the user’s mobile phone address book to prevent the anti-harassment software on the mobile phone from blocking voice calls and SMS notifications.

[0095] 4.1.5 Real-time feedback information

[0096] Feedback the user's call and hang-up actions and time points to the notification user interface in real time, so that the person sending the emergency notification can see it immediately, creating a sense of command and control with one call and a hundred responses. Figure 8 shown.

[0097] 4.2 Multi-cloud interconnection is convenient and controllable

[0098] 4.2.1 Convenient interconnection between multiple systems

[0099] This system is built on the cloud computing architecture, which is very convenient and simple to expand. Each bureau first builds its own local call system, and as long as the network is connected and the authorization is released, they can be directly connected.

[0100] 4.2.2 Reserve dedicated resources

[0101] Once a call channel is occupied, it cannot be interrupted. In order to avoid that the important emergency call requests of the own bureau cannot be responded to immediately after all the resources are occupied by calls from other bureaus, a certain number of call lines can be reserved for exclusive use by the bureau. For example, Guangzhou and Shanghai bureaus each built 200 call lines. In order to ensure the response speed of their own emergency services, each reserved 50 lines for their own exclusive use and did not share them with other bureaus.

[0102] 4.2.3 Good manageability

[0103] Unified management and operation and maintenance between multiple systems (multi-cloud) are carried out using Huawei's open source Kurator cluster management software.

[0104] The above description is only a preferred embodiment of the present invention and does not limit the present invention in other forms. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A highly flexible intelligent voice call system, characterized in that: Use Keepalived software to implement high availability features through the VRRP protocol, including: PBX calling system; JCWI Business Cloud Module; HSVCP-BE and HSVCP-FE are the front-end and back-end of the WEB management system.

2. The highly flexible intelligent voice calling system according to claim 1, characterized in that: The PBX call system is mainly responsible for fast text-to-speech conversion, service call notification, information collection, supports intelligent routing, call service intelligent queues, supports response level and number management, and can achieve multi-machine load balancing.

3. The highly flexible intelligent voice calling system according to claim 2, characterized in that: The JCWI business cloud module is mainly responsible for the communication between the business system and the voice call platform, as well as the shared services between clouds.

4. The highly flexible intelligent voice calling system according to claim 3, characterized in that: VRRP is Virtual Router Redundancy Protocol.

5. The highly flexible intelligent voice calling system according to claim 4, characterized in that: The PBX call system adopts a dual-path architecture of an application server and a call server.

6. The highly flexible intelligent voice calling system according to claim 5, characterized in that: The PBX calling system includes a voice calling platform and a background management system.

7. The highly flexible intelligent voice calling system according to claim 6, characterized in that: The voice call platform includes a business interface, JUSTTTS voice conversion, intelligent routing, JUSTIVR interactive voice response, underlying hardware interface and cloud functions.

8. The highly flexible intelligent voice calling system according to claim 7, characterized in that: The backend management system includes system management, application management, parameter configuration, statistical analysis, billing management and system monitoring.

9. The highly flexible intelligent voice calling system according to claim 8, characterized in that: The JCWI business cloud module adopts cloud computing architecture technology and is divided into two levels: first, the local system is built using a K8S application server cluster, and various network computing resources are used for load balancing and automatic fault transfer, which is equivalent to building a local cloud; second, the local clouds of each bureau are interconnected using Huawei Kurator for cluster management, making it easy to expand the cluster group. After the local cloud of a new bureau joins the cluster group, it immediately shares its own resources and simultaneously obtains the shared resources of other bureaus in the group.

Images