Unified deployment and multi-level management of WiFi network billing system in the context of a chain shop

The WiFi network billing system in the chain scene is particularly interesting: Headquarters wants to have a unified management, data and operating standards, but each store has its own network environment, bandwidth conditions and equipment type. A single-slash deployment sounds beautiful, almost always wally when actually landing. I saw a technological program for a local market WiFi project with headquarters standard of cloud-up management at one local billing gateway per shop, which is not a problem in itself, but when the first door is deployed, it becomes clear that the core exchange model in the store is different from what was assumed in the scheme, and VLAN configurations are all redesigned.

The first structural decision to run WiFi's network in a chain setting is whether it's centralized or distributed. Many programs are designed to be integrated into the system."Cloud tubes"The flag is actually all the authentication and billing logic on the cloud, with the doorman only keeping AP and simple network equipment. The benefits of this architecture are simple transport, data harmonization, and high quality requirements for the network links from the door to the cloud.——Once the doorman's Internet exports are out, not only WiFi is not usable but even locally available users cannot be properly certified.

The distribution structure is based on a local billing gateway in every door shop, which normally keeps pace with cloud management platforms, can run independently locally when the network breaks and then synchronizes data when the network resumes.——One hundred shops have 100 local nodes to maintain, upgrade and monitor.

The choice of which is not absolute or wrong is based on two indicators: how stable your door shop network is and whether your crew can support the day-to-day management of distributed deployments. If the door shop network is stable, the team has limited capacity, and a centralized approach is more appropriate; if the door shop network is not well organized but requires high business continuity requirements, distribution is the proper solution.

WiFi web-based billing systems require at least three layers of managerial authority: group management, regional management and door shop operations. The scope of terms of reference and operating capabilities vary from one level to another.

Group management can see data from all the shops, develop a global billing strategy and setup template, but not directly handle the daily business of a store.——Regional management can manage all the stores in their own region, and can set up regional marketing activities (e.g., a marketing package for East China), but cannot modify the overall strategy. The store operates only by looking at its own shop data, and only by operating its own doorkeepers and orders.

The system of access sounds simple, but it's two easily stepped pits. The first is data isolation.——A operator should not be able to see the user data for door shop B, which is not only a management issue but also a compliance requirement under GDPR or personal information protection laws. The second is strategy succession.——The corporate strategy for group settings can be covered by regional management? Can the store operating layer be overtaken by coverage? These rules of coverage are not well defined, and they must be messy when they go online.

WiFi web billing system in a chain of events has a special need.——Walking across the store. Can you log in automatically with the same account number when the same user is at a different door? A user who buys the package went to another shop belonging to the same brand, can it still be used?

This problem is common in hotel and mall chains. Technically, it can be achieved through a uniform identification and cruise paper.——User authentication information is stored in the central database, where local billing gateways verify user identification and package effectiveness to the central database every time they connect to WiFi at any store. But there is a commercial consideration here: branders may not want a package bought by users at Shop A to go to Shop B because the operating costs of different shops vary. This differentiated strategy requires flexible configuration capacity at the system level. The WiFi network accounting system project under the chain scenario is most feared that technology options are out of touch with management needs. The Headquarters technical team has a centralized approach to head-tapping, which results in inconsistent door shop network conditions that simply cannot be pushed; Or, in turn, to separate each store, headquarters would not pull out a full branding of business data. It would be better to take time to map the reality of the shops and then push back the technical architecture according to business objectives, rather than set the actual conditions for going to the shop instead. In practice, it is recommended that three or five different types of stores be selected for pilot deployment, using real operating data to verify whether the architecture is viable, and whether the whole brand can be extended. If brand names are a union model rather than a direct-account model, and the doorman is an independent ally instead of headquarters, then ownership of user data, the proportion of revenue allocated to billing, and whether there is authority at headquarters to access the door shop’s operating data are not only technical but also legal issues. These boundaries should be clearly defined with the legal and business teams during the design phase of deployment programs, and it should not be discovered that the conglomerates refuse to provide data after the system has been online.