Build your OWN Apple iBeacon with a Raspberry Pi

本文详细介绍了如何使用Raspberry Pi和蓝牙4.0 USB适配器构建一个类似Apple iBeacon的系统,包括安装必要的软件、配置Raspberry Pi以及解码iBeacon协议的过程。

Feature US department store Macy’s recently said it is implementing iPhone-based tracking tech the better to encourage browsing punters to buy. Of course, Macy has chosen to pitch this as an Apple technology - figuring, presumably, iPhone owners are more receptive to inducements delivered through technology and have more cash to splash than Android fans.

But the fact is, the system Apple calls iBeacon simply makes use of features already part of the Bluetooth Low Energy (LE) spec.

This got me thinking: how difficult would it be to build a similar system of my own? Not very hard at all, it turns out. Choose the right kit and it can be quite cheap too. I created my beacon using a £30 Raspberry Pi and a £12 Bluetooth 4.0 USB dongle.

Pimoroni Piglow

Surely this can’t be an Apple iBeacon? Yes it can

Bluetooth LE incorporates a protocol for beacon devices to identify themselves. Each sends out a short packet of data "advertising" which can contain up to 31 bytes of user-defined data. Apple’s iBeacon specification, such as it is, stores four values in this space: a “Proximity” 128-bit UUID and two 16-bit numbers, “Major” and “Minor”.

Apple has a good example of how these variables are used: a department store chain - Macy’s, say - adopts a single UUID for all its beacons. It uses the value of the Major variable to distinguish one shop from another, and the value of the Minor variable to differentiate between beacons in one shop’s departments.

Not all Bluetooth dongles are Linux-friendly. A handy resource listing well-behaving ones can be found at the Embedded Linux Wiki. A branded one will set you back around a tenner, generic ones less. I used IoGear’s GBU521.

Next, prepare your Pi. You need to install the official Linux Bluetooth software stack, BlueZ, and various USB development packages, some using the apt-get tool at the command line, others by compiling the code.

First run this:

sudo apt-get install libusb-dev libdbus-1-dev libglib2.0-dev libudev-dev libical-dev libreadline-dev

Next install BlueZ’s source files and compile it. The version at the time of writing was 5.11.

sudo wget www.kernel.org/pub/linux/bluetooth/bluez-5.11.tar.xz
sudo unxz bluez-5.11.tar.xz
sudo tar xvf bluez-5.11.tar
cd bluez-5.11
sudo ./configure --disable-systemd
sudo make
sudo make install

This will take a while, but when it’s done, you can reboot and plug in the dongle.

Decoding the iBeacon protocol

There’s no version of the uuidgen utility readily available for the Pi, so I used this website. The 16 pairs of two-digit hexadecimal values - each pair is dubbed an "octet" in the jargon - along with Major and Minor pair of octets, need to be punched into the Pi’s Bluetooth sub-system using BlueZ’s hcitool utility:

sudo hcitool -i hci0 cmd 0x08 0x0008 1E 02 01 1A 1A FF 4C 00 02 15 [ 92 77 83 0A B2 EB 49 0F A1 DD 7F E3 8C 49 2E DE ] [ 00 00 ] [ 00 00 ] C5 00

Note that the square brackets are NOT part of the command - I’ve added them solely to show where the UUID, Major then Minor codes go. The ‘C5’ after them is a value representing transmitted power level. Just cut and paste the line above and replace the UUID with your own.

Bluetooth 4.0 dongle

Not all Bluetooth dongles are Pi pals

This is how you decode the command: the "hci0" identifies your Bluetooth dongle, "cmd" tells hcitool to send the following command data to the device. The "0x08" is the Bluetooth command group - the "OGF" in the official parlance - and "0x0008" is the specific command ("OCF"), HCI_LE_Set_Advertising_Data.

The first "1E" is the number of “significant” octets in the advertising data that follow, up to a maximum of 31. The non-significant part should only comprise pairs of zeroes to take the number of octets up to 31 and which, to save power, are not transmitted.

The ad data is split into groups, each formatted with a single octet providing the number of remaining octets in the group - essentially it tells the Bluetooth sub-system how further along the list of octets is the next group. It’s followed by a single octet which defines the type of data, and then any number of octets holding the data itself. You can put as many of these groups into the advertising data packet as you can fit into the 31 octets allowed.

In my example, the first "02" in the sequence says the first block of ad data is two octets long. The next octet, "01" says the advertising octet(s) following are Bluetooth flags, and the "1A" is the binary value derived when certain of those flags are set.

‘1A’ says the next group is 26 octets long, and the "FF" identifies the group as manufacturer-specific data. The Bluetooth 4.0 specification says the next two octets have to expose the manufacturer: the "4C 00" is Apple’s Bluetooth manufacturer ID.

IBeacon notifications on Metawatch

In the Zone: location notification on a Metawatch smartwatch

I’m not yet sure what the "02" and "15" signify, but as I say, the Proximity UUID, Major and Minor values, and the power level complete the 26 octets of manufacturer data - and the 30 octets of the entire advertising data.

The hcitool command formats the iBeacon advertising signal. Telling the Pi to begin sending out that signal requires the following command:

sudo hciconfig hci0 leadv

You can disable LE beacon activity with the command:

sudo hciconfig hci0 noleadv

Update If you don’t see your beacon after issuing the leadv command, try sudo hciconfig hci0 noscan which stops the dongle looking for other Bluetooth devices. This can interfere with the beacon operation.

And it’s an obvious next step to create scripts to set all this up and activate LE advertising whenever the Pi boots up, but I won’t be covering that here. If you’d like to do that, there’s a very good tutorial written by Washington DC-based Radius Networks here.

Building a beacon monitor app

If you’d just like to cut to the chase and start coding iBeacon support into apps, Radius is selling pre-configured Pis as iBeacon development kits with one or two on-board dongles. Prices start at $100 (£61).

With the Pi running as a beacon, the next stage is to create an app that will look for it and notify you when you’re there. I chose to work up an iOS app - I’m exploring Apple’s iBeacon, after all - but it should be possible to code it up in Android 4.3, which added Bluetooth LE support to the Google OS.

Apple added it to iOS 7, released back in September. iOS’s existing CoreLocation framwork defines a CLLocationManager class that provides an interface for detecting iBeacons and a mechanism for dealing with events triggered by moving into and out a beacon’s zone of coverage, an extension of CLLocationManager’s already available ability to work with geographical regions.

CLLocationManager also defines methods for checking whether the device the app is running on has an OS and hardware able to handle Bluetooth LE:[CLLocationManager isMonitoringAvailableForClass:[CLBeaconRegion class]]returns a Boolean true if the device is beacon savvy.

Pass that test and a well-behaved app will then double-check that iOS’ Location Services have been enabled and that the app has permission to access them. It can then start looking for beacons with the Proximity UUID, Major and Minor values punched into the Pi earlier:

{
    ...
    
    CLBeaconRegion *beaconRegion;
    NSUUID *beaconUUID;
    NSString *beaconIdent;
    
    ...
    
    beaconUUID = [[NSUUID alloc] initWithUUIDString:@"9277830A-B2EB-490F-A1DD-7FE38C492EDE"];
    beaconIdent = @"Vulture.Zone.1";
    beaconRegion = [[CLBeaconRegion alloc] initWithProximityUUID:beaconUUID major:0 minor:0 identifier:beaconIdent];
        
    ...
        
    [appLocationManager startMonitoringForRegion:beaconRegion];
    
    ...
}

CoreLocation provides two handy delegate methods for dealing with events triggered by the beacon monitor: - (void)locationManager:(CLLocationManager *)manager didEnterRegion:(CLRegion *)region and - (void)locationManager:(CLLocationManager *)manager didExitRegion:(CLRegion *)region. The object that owns the CLLocationManager must adhere to the CLLocationManagerDelegateprotocol if it’s to receive these messages.

These two methods provide an easy way to alert the user he or she has entered the zone and to trigger actions accordingly. My test app, for instance, grabs a block of HTML code from El Reg’s server and presents it in a UIWebView: “Special Offer! The Editor will buy you a pint if you present this code...”, that kind of thing.

El Reg's iBeacon app

Running the app: We’re too far away (left) but then...

iOS scans for regions even when the app that initiated the monitoring isn’t running. The app is automatically run in the background if the beacon is detected in those circumstances. Likewise, it’s woken from sleep if it is merely napping. You can control whether messages triggered when beacon region boundaries are crossed are delayed while the iDevice’s screen is off: the CLBeaconRegion object created above has a Boolean property, notifyEntryStateOnDisplay, you can use to enable this behaviour. It also has a Boolean property called notifyOnEntry, inherited fromCLBeaconRegion’s superclass CLRegion, which you’ll need to set to NO in this case.

Once the phone knows it’s inside a beacon’s sphere of influence, the app can call the - (void)locationManager:(CLLocationManager *)manager didRangeBeacons:(NSArray *)beacons inRegion:(CLBeaconRegion *)region method to get a list of nearby beacons and from each entry, a beacon’s UUID, Major and Minor values to help identify which they are and, thanks to each beacon’s proximity property, roughly how close it is.

To keep things tidy, I added an switch to disable beacon scanning. By way of inter-object notifications, it calls [appLocationManager stopMonitoringForRegion:beaconRegion] to take the currently defined beacon off the system-wide monitoring list.

El Reg's iBeacon app

...and here’s a message from our (nearby) sponsor

Once all this is up and running, and the Pi-hosted iBeacon is operating in the background, it’s easy to test the system by walking out of range of the beacon and then turning round and coming back.

Android users keen to try this out without coding their own beacon detector can check out Radius Networks’ iBeacon Locate app in Google Play. There’s an iOS version in iTunes too. Radius also has an open-source library of iBeacon compatibility code for Android if you want to incorporate iBeacon support into an app of your own. ®


【Koopman】遍历论、动态模态分解和库普曼算子谱特性的计算研究(Matlab代码实现)内容概要:本文围绕【Koopman】遍历论、动态模态分解和库普曼算子谱特性的计算研究展开,重点介绍基于Matlab的代码实现方法。文章系统阐述了遍历理论的基本概念、动态模态分解(DMD)的数学原理及其与库普曼算子谱特性之间的内在联系,展示了如何通过数值计算手段分析非线性动力系统的演化行为。文中提供了完整的Matlab代码示例,涵盖数据驱动的模态分解、谱分析及可视化过程,帮助读者理解并复现相关算法。同时,文档还列举了多个相关的科研方向和技术应用场景,体现出该方法在复杂系统建模与分析中的广泛适用性。; 适合人群:具备一定动力系统、线性代数与数值分析基础,熟悉Matlab编程,从事控制理论、流体力学、信号处理或数据驱动建模等领域研究的研究生、博士生及科研人员。; 使用场景及目标:①深入理解库普曼算子理论及其在非线性系统分析中的应用;②掌握动态模态分解(DMD)算法的实现与优化;③应用于流体动力学、气候建模、生物系统、电力系统等领域的时空模态提取与预测;④支撑高水平论文复现与科研项目开发。; 阅读建议:建议读者结合Matlab代码逐段调试运行,对照理论推导加深理解;推荐参考文中提及的相关研究方向拓展应用场景;鼓励在实际数据上验证算法性能,并尝试改进与扩展算法功能。
本系统采用微信小程序作为前端交互界面,结合Spring Boot与Vue.js框架实现后端服务及管理后台的构建,形成一套完整的电子商务解决方案。该系统架构支持单一商户独立运营,亦兼容多商户入驻的平台模式,具备高度的灵活性与扩展性。 在技术实现上,后端以Java语言为核心,依托Spring Boot框架提供稳定的业务逻辑处理与数据接口服务;管理后台采用Vue.js进行开发,实现了直观高效的操作界面;前端微信小程序则为用户提供了便捷的移动端购物体验。整套系统各模块间紧密协作,功能链路完整闭环,已通过严格测试与优化,符合商业应用的标准要求。 系统设计注重业务场景的全面覆盖,不仅包含商品展示、交易流程、订单处理等核心电商功能,还集成了会员管理、营销工具、数据统计等辅助模块,能够满足不同规模商户的日常运营需求。其多店铺支持机制允许平台方对入驻商户进行统一管理,同时保障各店铺在品牌展示、商品销售及客户服务方面的独立运作空间。 该解决方案强调代码结构的规范性与可维护性,遵循企业级开发标准,确保了系统的长期稳定运行与后续功能迭代的可行性。整体而言,这是一套技术选型成熟、架构清晰、功能完备且可直接投入商用的电商平台系统。 资源来源于网络分享,仅用于学习交流使用,请勿用于商业,如有侵权请联系我删除!
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值