I was reading QSDK page [1] and it seems to me that they are doing something different from openwrt.
>The QCA Software Development Kit (QSDK) project allows users to build an OpenWrt based platform containing additional enhancements for Qualcomm Atheros chipsets that have not yet made it into the public OpenWrt repository.
I'm not aware of the project goals so I may be wrong.
Yeah. It was a very good read. I could never have thought that you could use xmpp protocol for microblogging! Movim's the only platform that does that as far as I can tell.
This is a great idea, well done! I think as a next step you should look a bit more into routers themselves. You can flash them with an opensource os like openwrt [1] and run a chat (or some kind of communication) server on it so everybody connected to the same wifi can freely exchange data.
This way you can turn any wireless capable [2] machine into a small community network.
> great use case I've seen are college campuses where thousands of students connect to the same WiFi,
This is indeed a great use case. Right now you're only thinking about chat but its got more applications like semi-remote teaching,tests ,role calls etc. But (I think) its far easier to realize it if you work at the network level rather than at the app level.
One of the things that I find really encouraging is governments letting other industires to become ISP
>In Tennessee, for example, state laws allow publicly-owned electric utilities to provide broadband
I think this is the best way to solve last mile distribution. Electric companies already have the capacity to manage door to door to installations. Internet could become an additional source of revenue for them. They could lease the line from big providers and increase their reach. Win Win.
Fatalism works in both directions. If you believe that your destiny is fixed and no matter what you do is going to make a big difference then you can do anything you want content in the knowledge that nothing can stop you for achieving your destiny. Once you realize this you'll understand that fatalism as a philosophy defeats itself. There a very good essay by Emerson on this [1].
Additionally I would like to add, read everything but don't get drowned in the teaching of idle pessimistic philosophers (who they are is up to you to decide I have my list make your own). They are rightly derided for having nothing better to do than to preach. Learn to punch holes in a theory. You can start by punching a big one in mine.
>For, if Fate is so prevailing, man also is part of it, and can confront fate with fate. If the Universe have these savage accidents, our atoms are as savage in resistance. We should be crushed by the atmosphere, but for the reaction of the air within the body. A tube made of a film of glass can resist the shock of the ocean, if filled with the same water. If there be omnipotence in the stroke, there is omnipotence of recoil.
I'm interested in building commission free marketplaces. Some of the advantages that a system like this can have is
- You can completely get rid of centralized international payment processing and distribution.
- Countries where cellular networks have good coverage (i.e. most of the regions where marketplaces operate today) already have highly efficient payment networks that can be set up by any individual or a business. All we need to do is educate them how.
I don't think building a technology infrastructure would be the right way to solve this problem. Effort should instead be spent in educating how the existing infrastructure can be used by individuals to their advantage.
The goal of marketplace operator should be to ensure that all the operations are run smoothly. How customer grievances are to be addressed? How the legal regulations are to be followed? How the service is going to be provided? How the compensation is going to be received? In other words the marketplaces should take away the administrative burden of the participants.
The wifi wave protocol or 802.11p [1] has been the cause of recent contest between the EU parliament and EU Telecommunication Network Operators association (ETNO). The dispute being that a wifi only standard will stall the growth in the connected car segment.
The misgivings of ENTO are well founded.
> "In fact, research also shows that customers are willing to switch manufacturers just to be able to use mobile devices and connectivity. In 2014 there were 21% who were willing to do so whereas in 2015 this number climbed up to 37%." [2]
As the research shows people want better mobile connectivity inside their cars. Since 5g is a new standard with few real world deployments access points inside the cars would make it a very attractive package. However there are numerous hurdles
1. 5g deployments are expensive. New equipment needs to be manufactured to create embedded access points inside cars.
2. 5G is a licensed spectrum. If the technology is going to be used inside the cars automakers will need to apply for a licence. Or work with licence acquirers. (It will not be too far fetched to imagine for example a Vodafone car in the future) However this does put small vehicle manufactures at a very big disadvantage for a long time.
3. 5G is a cell phone network. It is deigned and optimized for use in mobile phones (with some afterthought to IOT). Aside from the fact that a car is not a mobile phone who is going to bear the cost of data? Or put in other words why does a connected car network need a telephone network? It is understandable that the telephone operators want to put a sim card in any electronic device they can to increase their revenue but should there be not a comparable benefit to the subscriber of the network? Should a car owner pay the phone bill for his car as well? Do car manufacturers want to build car-as-a-service?
This brings me to ask
What makes any standard effective?
The implementation of the standard. HTTP is an effective standard not because it a protocol defined in research papers but it has been adopted by the community.
It has not been adopted by the community because it is a protocol defined in a paper but because client applications have implemented it.
The client applications implemented HTTP not because it was the invention of smart folks who wrote effective research papers but because it would be easier to gain the support of the community to build applications for the client.
What is the most effective wireless standard in use today?
It is wifi. It is available everywhere. Even in $7 feature phones. It is also free of cost.
What are the deficiencies of wifi in the connected car scenario?
No information car to car exchange can take place unless
- Either both the devices are connected to the same base station
- Or The devices are paired with each other.
Wifi direct seems to solve this problem but it is not well implemented in client devices. And even if the hardware support it new software will have to be written to get it to work as expected.
How can we work with these shortcomings?
Easily. While direct p2p communication is difficult on a physical layer. Several hundred application layer services exist that enable p2p communication. Any linux device with a network card can run a p2p service for efficient communication.
Do we need 5g in cars?
>If we mandate only an older technology like Wi-Fi 802.11p and exclude a future-proof one like 5G, we put Europe behind global peers like the U.S., South Korea and China.” [1]
A G network can hardly be called future proof. Over the decades there have been 5 revisions already with each revision requiring a device upgrade from the client.
Wifi on the other hand has been more stable. It is a well understood technology that works everywhere. All wifi enabled device work the same way. A lot of knowledge has been built on how to operate and configure a wifi network.
Additionally wifi networks don't require a skilled worker. Even novice users understand and are able to use it.
Despite all the advantages of 5g, Wifi is the true technology and vendor neutral solution for all kinds of connected infrastructure technologies.
Yep I think this is a very good idea too. It is a way of reusing existing spectrum space. Besides small cell networks use low powers and being portable they can deployed on demand.
In fact small cell networks might be how the networks are deployed in future. Website below has some interesting research on this topic
You are correct. Most router these days select channels automatically.
But the automatic channel selection is not very accurate in my personal observation as well as that of a few others on the internet [1]. Wish it were better though. I feel that there are numerous problems in automating the channel selection
1. We have to make sure that this happens without any downtime otherwise there is poor QOS. If the router has to reboot to choose a channel you're out of internet for about 5-10 minutes. It can be very frustrating if it happens often.
2. As if that was not enough you have to draw extra power to scan nearby access points and constantly react to the changing signals. (This would also defeat the point of using low power signals)
In my opinion it is better to tune the routers manually than to set it on automatic channel selection.
> the odds that your [...] is _very_ low
Yes unfortunately that is true. There is little we can do about it. But there is one redeeming factor. As per my understanding interference depends not only on the base access point but also on the client. In other words the channels will deteriorate if the communication link that is the pairing of client with the access point interferes with the signals of another access point.
On low ranges you can create mini line of sight networks that will happily co exist with any number of signals. Example wifi on laptop working with bluetooth mouse.
[Router] --------- [laptop]
|
|
[mouse]
It is not such a good example because signals at such a small distance are strong enough to overcome any interference but the principle stands in other conditions as well.
2.4ghz also has 3 non overlapping channels (iirc). Most people use the default channel of their WiFi so if you choose a different channel you can get a better signal at a lower power.
Edit
If you want low coverage area it would also not be such a bad idea to use the 5ghz frequency if your router supports it. Much less crowded shorter range by default and higher speeds
I'm not really aware of how FCC collected this data and how the analysis were done. I'm also not aware of the policies that are set by the US government to provide rural connectivity. But I have a few general thoughts on the issue.
Estimating the broadband adoption is a tricky issue because of number of stakeholders involved. First there are the companies that provide the physical connectivity. Their job is done as soon as the lines are laid to the base station or the satellite is sent into the space, etc [1] According to them anywhere the physical layer has reached there is internet connectivity (which is true...to an extent)
But this does not actually get the internet in to your home. Local distribution is a very big challenge. A similar thing happened in India with Bharat net [2]. While the government successfully laid down the fiber optic cables the last mile connectivity was overlooked. So of course the broadband is inadequate when the capacity is not being used.
I agree with the article that internet is connected to a higher economic growth and overall prosperity. The same sentiment is also echoed by the WEF report [3]. However these reports miss one very crucial point. Internet is not a service provider. Internet is a transport. You need to have people provide service on the internet if it is to be used. Local internet economy is almost non existent.
Many of the faults in FCC's report can be explained if you look at the internet provider business as a marketplace.
- Big companies own the platform and assert that the platform can reach anywhere. Which might very well be true.
- Local internet service providers use this platform to provide services to the customers.
- Customers pay for the internet service according to usage.
Why should a customer pay heavy charge for the broadband usage? Why should the local internet service provider invest into last mile connectivity?
On a global scale internet rapidly reached new levels of adoption when the application layer succeeded. Something similar needs to happen on a local level, I think. And this is what microsoft is also trying to do with their airband platform [4].
[1] If you do a bit of digging into physical layer distribution there are many exotic ways to provide network connectivity including sending weather balloons up in the air! Very interesting area of research if you're into this sort of thing.
Yes. Telephone companies can do this. You are correct about it.
I don't have an answer to your question so I'm just speculating over here and it's quite possible that I'm wrong
I think as soon as that happens (someone shuts off a network) someone else will set up new networks to provide the old service.
Open source technologies exist to set up base stations right now [1]
I think we're headed towards an open spectrum future. Soon most communication could take place on unliscenced networks [2]. Or more and more networks could become unliscenced.
The G networks are not the only way to communicate. It is also important to note that the G networks are now simply internet enablers.
I don't have any data on this but I'd love a comparision of telephony and data services on use today. How much data people use in calling vs using Facebook for example?
Also consider that all telephony comapnies are betting big on IOT.How many IOT devices works on a G network?
If you have a modern device all you need is WiFi.
That however does not mean that telephone companies would cease to exist. Someone still has to provide infrastructure. They just won't have to spend so much on acquiring liscence s. Which also means more competition from unliscenced providers.
Device manufacturers would be able to work on a common network technology stack.
By the way I don't have anything against 5g. I also don't quite understand health implications so I'm unqalified to talk about it. But I don't like having to buy a new device just to use a new network.
> 4G, just like 3G before it will get slower over time.
You are correct but that will happen again with 5G and then with 6G and so on. After a period all networks will saturate.
I'm not sure if I'm right about this, but I think one solution could be to increase the network density with the deployment of small cell networks that provide a better local coverage and bandwitdh. A few companies like nokia [1] and ericsson [2] are taking that route.That also requires upgradation on hardware form the operator but at least spares the customer from buying a new equipment.
>The QCA Software Development Kit (QSDK) project allows users to build an OpenWrt based platform containing additional enhancements for Qualcomm Atheros chipsets that have not yet made it into the public OpenWrt repository.
I'm not aware of the project goals so I may be wrong.
[1] https://wiki.codeaurora.org/xwiki/bin/QSDK/