I have a BLE peripheral running on Raspberry Pi with gobbledegook.
In parallel, I have an iOS app working as BLE central which does a scanForPeripheralsWithServices with nil in order to detect all the devices around.
I'm able to detect a few devices, but not the one running on Raspberry Pi.
However, when scanning with the nRF Connect app on iOS, I'm able to see it.
How is that possible? (I assume that nRF Connect is also using Core Bluetooth...)
You're likely doing something (such as using the nRF Connect app) to connect to the device, causing it to stop advertising. Most BLE platforms stop advertising when they are connected to.
Use retrieveConnectedPeripherals(withServices:) to detect devices that are already connected. You must provide a list of services; there is no way to detect all devices currently connected.
Since this is on a Pi, you can also verify on that side whether the system is really currently advertising and whether it has any connections. Core Bluetooth does extensive caching, which can lead to things appearing to be happen that actually are coming out of cache. I haven't recently dug into the nRF app to verify how much it relies on the cache. I recommend also testing with LightBlue, and testing with multiple devices simultaneously (I typically use multiple apps on iOS, Android, and Mac, and occasionally an Ubertooth, to verify what I'm seeing is true).
Note also that once you've seen an advertisement from a device once, you won't see any more during that scan unless you set CBCentralManagerScanOptionAllowDuplicatesKey to true in you scan options. That flag is ignored in the background. This probably isn't the issue in this case, but often trips people up who are used to other platforms. Similarly it is almost always faster and more reliable to scan for specific services rather than nil unless you're building a generic BLE scanner. (Again, I doubt that's the problem in this case.)
Related
I have been fiddling with Bluetooth lately and I feel as if there is some elusive thing I am missing, anyways here are some of my questions. What I want is to have my Bluetooth enabled laptop running Windows, communicate with my iOS and later Android devices.
Is Bluetooth LE compatible with older Bluetooth technologies?
-My believe is that Bluetooth LE is the same as other Bluetooth versions
only that of course, as the name implies, it is low power. My confusion arises from the fact that on the iOS page they refer to some of the schemes as GAT servers , and what I am guessing is another protocol, ATT. Upon reading the article posted Here there is no mention of these two only L2CAP. As well as there is no mention of those in the Bluetooth API for windows Programming with Windows sockets The protocol that this article mentions is RFCOMM
-How is a connection between the two devices started?
I know that there are service ID's which can be published in an internal Service Description server according to the article previously referenced. However there is no concrete mentioning on the iOS bluetooth pages. They mention peripherals advertising bundles of data and the central device connecting to it. No mention of SDP, unsure if the windows docs do. And I' am bit confused about how the process flows, do the two devices connect first and then one device checks what services are available and the other one subscribes? Here by services I mean which applications are running, and by connection I mean as in establishing a connection from say (in iOS) Settings>Bluetooth and then starting the application on your phone, and then it is able to scan, or can the process of enabling Bluetooth and establishing a connection be done in the application itself. Same for windows, do I have to start the Bluetooth device by Settings>Bluetooth or can it be triggered within the Bluetooth enabled application?
-Is it pivotal to assign which device is the central and which is the peripheral?
Since at the end of the day they are able to both exchange data. In my case I want to have my computer running windows communicate with my iOS 7 device. Now as mentioned there are two roles that you could choose from in iOS: Peripheral and Central. It makes sense to me that the phone should be a peripheral since the main application will run on my computer, and I want my computer to enter the listening state before the device, and have the device connect afterwards. But it makes sense to have my computer advertise its services and then letting the iOS application choose the correct one. If someone could give me some pointers on what the best practices are and how to tackle this problem, I'd appreciate it.
I'm developing an application on iOS and OS X Yosemite which is using CoreBluetooth on iOS and IOBluetooth on Mac.
The iOS app plays the "central" role while the Mac app is playing the "peripheral" role.
Basically, the iOS app (the central) starts a scan to find peripherals and when it finds a Mac advertising peripherals data, the central connects to it and can discover its services and characteristics.
My problem is that sometimes, I don't know why, the iOS app starts a scan to find a Mac advertising and when it finds it (this step work all the time), it tries to connect to it but never succeed.
To make it able to connect successfully to the peripheral again, I need to turn off and then turn on the Bluetooth from the iOS button.
Another strange thing is : imagine the central is in the same situation (trying to connect with no success). On OS X if I click on the Network icon (with the Wifi symbol) in the system status bar, then it will display the list of all available Wifi networks, and when it adds to the list my iPhone as a "Personal hotspot", at this time the central (which is on the iPhone) manages to connect to the peripheral (the Mac).
With this problem, the app usage is a lot compromised because randomly the iPhone will not be able to connect to the Mac.
I spent a lot of days on the internet (Google, Stackoverflow, ...) to find similar cases. I read a lot of articles and tried a lot of things like stopping the scan and starting it again if the connection didn't succeed after 10 seconds for example.
Don't hesitate to tell me if my question is not clear. I will update it if needed.
Thanks a lot in advance for your help.
I really need to find a way to fix this.
I don't have an answer (and not enough rep to comment) but I have been experiencing a very similar problem for months now. My iOS app which is playing the central role sometimes fails to discover services on the peripheral (linux) which can only be fixed by power cycling the bluetooth radio on my iOS device via settings/bluetooth. After bouncing the bluetooth radio, peripheral discovery appears to works perfectly.
I've found two reproducible cases: (1) If on my peripheral I completely swap out the service being advertised (i.e. change the service UUID and all characteristics) then the iOS peripheral will detect the peripheral with this new service during the scan, but service discovery fails. I find that peripheral:didDiscoverServices gets invoked, but the list of services is empty for this peripheral. (2) If on my peripheral, I modify the list of characteristics in any way (add a new one, remove one) the iOS app does not notice this change at all. During characteristics discovery, the same list of characteristics (prior to modification) is presented. Again, power-cycling the bluetooth radio on the iOS device appears to be the only thing that fixes this.
It appears as though iOS is caching BTLE peripheral services and characteristics upon discovering a new peripheral and the only way to flush this cache is to power cycle the bluetooth radio.
I am working with hardware that is in the process of being developed. The board uses a TI CC2541 as the bluetooth chip.
Prior to a test the board is discoverable by two devices and I am able to connect to the board and perform a test where I write and read characteristics (GATT). I perform the test again, everything goes fine until I reconnect to the device and attempt to discover its services. At that point, the board is offline and neither lightblue nor my own app can detect the device. I did a factory reset on one of the iOS devices (iPads) and it is still unable to recognize the board.
Should a factory reset remove all corebluetooth caching that has happened on the device, allowing iOS to re-cache and rediscover the board?
Can this possibly be a corebluetooth problem or a problem with my app?
Your problem seems to be related to the external peripheral. Factory reset will remove any cached data from the iOS device so as long as the peripheral behaves correctly, the rediscovery should happen as planned. This is all I can tell from the info you shared.
This issue seems to be related to hanging / adding characteristics on a peripheral during development.
I have had a similar issue and instead of clearing the cache on the iOS device, I change the Bluetooth address of the peripheral. This allows the discovery to work as expected.
I am having an issue concerning Bluetooth Low Energy. I am using an iPhone (iOS 6.1) and an Nordic board using the nRF51822 chip.
I am currently developing a system where the iPhone and the nRF51822 connect using certain services, then disconnect and reconnect again. However, on the second connection, the nRF51822 uses different services from the first connection.
The problem is that on the second connection, the iPhone discovers the former services of the first connection. I have tested with a TI CC2540 development board and it discovers the new services well. The only solution I have found is to put the central manager to nil and allocate it again to make it work. Nevertheless, it seems kind of a dirty solution, is there a standard (clean) way to deal with this kind of situation?
If you don't do any bonding, but only pairing, then you would not have this issue.
If you do require bonding, then the best way to solve this problem is to use the Bond-manager that is in the SDK. If you are not using the bond manager, you could try to use the Service Changed functionallity and set the start handle to 0x000C, and end handle to 0xFFFF. This would most likely trigger a rediscovery from the iOS, thus finding your new services and characteristics. The iOS would always look for service changed, and enable this in the peripheral.
If you have any Nordic related question, then you could post them at the Q&A forum located at https://devzone.nordicsemi.com/
BR
Pål
Services and characteristics are cached on the iOS side when you initiate a pairing request with a ble device and discover services/characteristics. Don't do the pairing request if you're concerned about that. Then just pass in the array of services you actually want to discover (not nil).
(Side note: The whole thing of having the ble device controlling which services are seen to the iOS device seems really muddled..you should control that on the iOS side.)
I am working on a major product release for my company. We are designing a new device that we would like to integrate into an iOS app. (FYI, we have recently been accepted into the Apple MFi program so we are past that step.)
What we desire is to create a hardware/software eco-system where our users can start our iPad app (yes, it is specific to the iPad and the not iPhone or iPod Touch devices) and the application will automatically discover and link up with any close-proximity hardware that we've developed.
This means that I will need to implement some form of device connection process on both the iPad and within the hardware.
Is there any way to code automatic connectivity to a Bluetooth device from within an iOS application?
I am aware of the Core Bluetooth Framework which has this capability, however, it only works with the new iPhone 4S. That's not an option.
I've also researched about the External Accessory Framework, which can be used to connect to external devices whether they are connected via the iPod port or Bluetooth. This is an option but, if my understanding is correct, our hardware must already be paired with iPad before the iOS software can connect to it. This is less than preferable because although many of our devices may not be in immediate vicinity at one point in time, our customers could potentially own 100's devices that they may desire to connect to with our software.
What are my reasonable options or alternatives? The end goal is to provide a very easy means for our users to be able to connect to our devices, from within our software, on an iPad.
NOTE: I MUST NOT jailbreak the iPad or break any Apple TOS agreements.
UPDATE (3-7-2012) I saw today that the 'New iPad' (as it is being officially called) will be Bluetooth 4.0 capable. I assume this means that this iPad device will be able to use the Core Bluetooth Framework. This doesn't immediately solve my problem because we do want remain compatible with at least the "last generation" iPad device, but this is worth pointing out for others who may be looking for such an answer.
You cannot connect an iOS device to a bluetooth device without pairing first. It is purposely built that way by Apple to reduce the likelihood that someone could connect a bluetooth device without the user knowing about it. If pairing each device is unacceptable (and it sounds like it is), and you don't want to develop a dongle, the only remaining path I can see is to make your devices capable of relaying data between them like a mesh network. That way the user would only have to pair with a single device, and that paired unit would relay communications between the other devices and the iPad.
One of your alternatives would be going WiFi (AdHoc or Infrastructure) with Bonjour. For example you would be able to detect the device nearby (if it is maintaining it's own AdHoc then by SSID; if it's in infrastructure network then by service being announced). One big drawback would be that you couldn't connect directly (from app) to the device's network - user would have to do it manually.
As for BT: big drawback is lack of a serial protocol (not implemented in iPhone/iPod/iPad).