-
Amazon Deals - ToS - Like us?
National Broadband Plan, TRAI and the Indian broadband customer
- Thread starter itzmynet
- Start date
- Replies 151
- Views 45,633
mgcarley
Founder, Hayai Broadband
For faster Internet, embrace Communism.
Nah, most of the communist countries have shit Internet. For good Internet, you want socialism.
That aside, I think it would be interesting to try and procure bandwidth from China through Nepal or Sikkim, because I know China has a lot of cables from it's east coast to Japan and the US, so it might make US routes much cheaper (if technically feasible to get them in to India, of course)... though I suspect the Indian government would not like that for security reasons.
If we did that, it might also give us a place to terminate VOIP calls as well (of course, we still can't legally assign phone numbers with the Indian country code), but most VOIP wholesalers are selling China and India routes at very similar prices.
Nah, most of the communist countries have shit Internet. For good Internet, you want socialism.
That aside, I think it would be interesting to try and procure bandwidth from China through Nepal or Sikkim, because I know China has a lot of cables from it's east coast to Japan and the US, so it might make US routes much cheaper (if technically feasible to get them in to India, of course)... though I suspect the Indian government would not like that for security reasons.
If we did that, it might also give us a place to terminate VOIP calls as well (of course, we still can't legally assign phone numbers with the Indian country code), but most VOIP wholesalers are selling China and India routes at very similar prices.
Shreeraman
Newbie
Nah, most of the communist countries have shit Internet. For good Internet, you want socialism.
That aside, I think it would be interesting to try and procure bandwidth from China through Nepal or Sikkim, because I know China has a lot of cables from it's east coast to Japan and the US, so it might make US routes much cheaper (if technically feasible to get them in to India, of course)... though I suspect the Indian government would not like that for security reasons.
If we did that, it might also give us a place to terminate VOIP calls as well (of course, we still can't legally assign phone numbers with the Indian country code), but most VOIP wholesalers are selling China and India routes at very similar prices.
Keeping the security thing aside.... many Indian telecos has earlier said they plan to link China and India over land... through the harsh Himalayan terrain...! ISP's like Airtel, TATA-VSNL.... some links are here.
If the government keep on mentioning security... as a hurdle... then ISP market will be a real graveyard in India.
Shreenesh Raman
That aside, I think it would be interesting to try and procure bandwidth from China through Nepal or Sikkim, because I know China has a lot of cables from it's east coast to Japan and the US, so it might make US routes much cheaper (if technically feasible to get them in to India, of course)... though I suspect the Indian government would not like that for security reasons.
If we did that, it might also give us a place to terminate VOIP calls as well (of course, we still can't legally assign phone numbers with the Indian country code), but most VOIP wholesalers are selling China and India routes at very similar prices.
Keeping the security thing aside.... many Indian telecos has earlier said they plan to link China and India over land... through the harsh Himalayan terrain...! ISP's like Airtel, TATA-VSNL.... some links are here.
If the government keep on mentioning security... as a hurdle... then ISP market will be a real graveyard in India.
Shreenesh Raman
mgcarley
Founder, Hayai Broadband
Keeping the security thing aside.... many Indian telecos has earlier said they plan to link China and India over land... through the harsh Himalayan terrain...! ISP's like Airtel, TATA-VSNL.... some links are here.
If the government keep on mentioning security... as a hurdle... then ISP market will be a real graveyard in India.
Shreenesh Raman
It appears the Chinese are facing regulatory issues as well... why am I not surprised?
They've basically dismissed having a JV or anything in India, but overall I'd say the land-cable could be a good idea - so long as the Chinese didn't perform any filtering or monitoring on traffic passing through their country (as Sweden does with all traffic flowing through that country, whether it's bound for a Swedish server/user or not).
I think that a China-India cable would open up the market in such a way that ISPs would be able to buy US-routes in China and send them to India over the cable from China which may work out cheaper depending on Chinese domestic transit costs to get from the east coast of China to the China-India border. If we had the required cash (probably a few hundred million dollars), I'd build such a cable in a heartbeat. Tata has actually been talking about a China-India terrestrial cable since 2009 but so far nothing seems to have happened.
End of the day, I feel India-Asia and India-Europe route prices need to fall dramatically - probably 75 to 95% at least - to between Rs300 and Rs500 per mbit/s at wholesale - in order to fuel a true internet revolution in India. When that happens, I feel it would be feasible to provide something like up to 20mbit/s on DSL or 100mbit/s on fiber for Rs2k-2.5k without worrying about any FUPs and such - even if everything else stayed equal.
I personally feel that the wholesalers would sell a pile more bandwidth than they do now if they did that, too - instead of buying just 1-5Gbit/s or whatever as the smaller guys do now, ISPs might be inclined to buy 10-50Gbit/s instead and just pay the same amount of money: I know that's what I would do.
And it is affordable for them to do that - let's say it costs 1 Billion dollars to build a cable from Singapore to London, with 2 redundant fiber pairs, each with a minimum capacity of 10gbit/s per wavelength and 64 wavelengths per pair, thus equalling 1280Gbit/s or 1.28Tbit/s (that's a pretty low capacity for a new cable but what the hell, for the sake of an argument, let's use these figures even if the costs are slightly inflated and the capacity is slightly less than what I'd expect for a cable costing that much...)
$1 billion works out to $1,280,000 per Gigabit over the life of the cable.
Which works out to $1,280 per Megabit over the life of the cable.
Which works out to $10.67 per month per megabit for the entire length of the cable (assuming 10 Year ROI - although cables are generally expected to have a service life of 20-25 years).
Let's say if we add some maintainence costs and perhaps interest on any loans and a bit of profit, the cost to the ISPs becomes then $15 per month per megabit with larger volumes and say $30 per month per megabit for lower volumes - this works out somewhat favourably to all involved, because essentially we have to pay twice because we end up purchasing 2 routes out of India - one to Singapore and one to Europe.
If we end up paying Rs300 and Rs500 for each destination respectively, this falls in to the price ranges I'm talking about - about $18 per mbit over the whole span of the cable (though of course, we only buy which routes we need, so if we buy exclusively India-Singapore routes then it only works out to less than $7 per mbit). Of course, operating on this premise assumes that the cable would operate at quite high percentage of capacity from day-1, but the idea here is that with lower wholesale pricing, it would not be difficult to attract either larger volumes or more buyers.
But the current charges start at well over $120 per month per megabit to Singapore at the moment (costs are even more to Europe since the cost is determined by distance, so for the entire span of the cable we might end up paying something like $300/mbit for the entire span of the cable), which suggests that even at about 35% capacity on the main cable, SMW4, the wholesalers are making a killing, and I would suggest that is roughly half the reason for most ISPs implementing FUPs (not to mention the last mile costs/difficulties and peering ecosystem etc), and half the reason that only about 4% of the total capacity of India's submarine cables are lit up at the moment.
Out of about 23Tbit/s, less than 1Tbit/s are currently lit across all ISPs - Airtel has about 100Gbit/s, Tata is not clear for it's India operations (but over 1Tbit/s for it's US operations), Reliance has about 100Gbit/s, BSNL does not have any publicly available information but I'd suggest it's consuming around 300Gbit/s based on the number of IP addresses it has.
If the government keep on mentioning security... as a hurdle... then ISP market will be a real graveyard in India.
Shreenesh Raman
It appears the Chinese are facing regulatory issues as well... why am I not surprised?
They've basically dismissed having a JV or anything in India, but overall I'd say the land-cable could be a good idea - so long as the Chinese didn't perform any filtering or monitoring on traffic passing through their country (as Sweden does with all traffic flowing through that country, whether it's bound for a Swedish server/user or not).
I think that a China-India cable would open up the market in such a way that ISPs would be able to buy US-routes in China and send them to India over the cable from China which may work out cheaper depending on Chinese domestic transit costs to get from the east coast of China to the China-India border. If we had the required cash (probably a few hundred million dollars), I'd build such a cable in a heartbeat. Tata has actually been talking about a China-India terrestrial cable since 2009 but so far nothing seems to have happened.
End of the day, I feel India-Asia and India-Europe route prices need to fall dramatically - probably 75 to 95% at least - to between Rs300 and Rs500 per mbit/s at wholesale - in order to fuel a true internet revolution in India. When that happens, I feel it would be feasible to provide something like up to 20mbit/s on DSL or 100mbit/s on fiber for Rs2k-2.5k without worrying about any FUPs and such - even if everything else stayed equal.
I personally feel that the wholesalers would sell a pile more bandwidth than they do now if they did that, too - instead of buying just 1-5Gbit/s or whatever as the smaller guys do now, ISPs might be inclined to buy 10-50Gbit/s instead and just pay the same amount of money: I know that's what I would do.
And it is affordable for them to do that - let's say it costs 1 Billion dollars to build a cable from Singapore to London, with 2 redundant fiber pairs, each with a minimum capacity of 10gbit/s per wavelength and 64 wavelengths per pair, thus equalling 1280Gbit/s or 1.28Tbit/s (that's a pretty low capacity for a new cable but what the hell, for the sake of an argument, let's use these figures even if the costs are slightly inflated and the capacity is slightly less than what I'd expect for a cable costing that much...)
$1 billion works out to $1,280,000 per Gigabit over the life of the cable.
Which works out to $1,280 per Megabit over the life of the cable.
Which works out to $10.67 per month per megabit for the entire length of the cable (assuming 10 Year ROI - although cables are generally expected to have a service life of 20-25 years).
Let's say if we add some maintainence costs and perhaps interest on any loans and a bit of profit, the cost to the ISPs becomes then $15 per month per megabit with larger volumes and say $30 per month per megabit for lower volumes - this works out somewhat favourably to all involved, because essentially we have to pay twice because we end up purchasing 2 routes out of India - one to Singapore and one to Europe.
If we end up paying Rs300 and Rs500 for each destination respectively, this falls in to the price ranges I'm talking about - about $18 per mbit over the whole span of the cable (though of course, we only buy which routes we need, so if we buy exclusively India-Singapore routes then it only works out to less than $7 per mbit). Of course, operating on this premise assumes that the cable would operate at quite high percentage of capacity from day-1, but the idea here is that with lower wholesale pricing, it would not be difficult to attract either larger volumes or more buyers.
But the current charges start at well over $120 per month per megabit to Singapore at the moment (costs are even more to Europe since the cost is determined by distance, so for the entire span of the cable we might end up paying something like $300/mbit for the entire span of the cable), which suggests that even at about 35% capacity on the main cable, SMW4, the wholesalers are making a killing, and I would suggest that is roughly half the reason for most ISPs implementing FUPs (not to mention the last mile costs/difficulties and peering ecosystem etc), and half the reason that only about 4% of the total capacity of India's submarine cables are lit up at the moment.
Out of about 23Tbit/s, less than 1Tbit/s are currently lit across all ISPs - Airtel has about 100Gbit/s, Tata is not clear for it's India operations (but over 1Tbit/s for it's US operations), Reliance has about 100Gbit/s, BSNL does not have any publicly available information but I'd suggest it's consuming around 300Gbit/s based on the number of IP addresses it has.
Out of about 23Tbit/s, less than 1Tbit/s are currently lit across all ISPs - Airtel has about 100Gbit/s, Tata is not clear for it's India operations (but over 1Tbit/s for it's US operations), Reliance has about 100Gbit/s, BSNL does not have any publicly available information but I'd suggest it's consuming around 300Gbit/s based on the number of IP addresses it has.
IMO, all Indian ISPs are EVIL. What do you have to say about this? I still don't understand. When you have the 100% capacity, why just make use of 1% and in turn impose BS Unfair Usage Policies like what Airtel is currently doing? I just don't get it. Why not light up more and introduce some sane broadband tariff plans? Why is India still stuck in the stone age when it comes to broadband? IMO, new new 512 Kbps broadband definition ain't gonna change anything.
IMO, all Indian ISPs are EVIL. What do you have to say about this? I still don't understand. When you have the 100% capacity, why just make use of 1% and in turn impose BS Unfair Usage Policies like what Airtel is currently doing? I just don't get it. Why not light up more and introduce some sane broadband tariff plans? Why is India still stuck in the stone age when it comes to broadband? IMO, new new 512 Kbps broadband definition ain't gonna change anything.
- Messages
- 493
- Location
- NA
IMO, all Indian ISPs are EVIL. What do you have to say about this? I still don't understand. When you have the 100% capacity, why just make use of 1% and in turn impose BS Unfair Usage Policies like what Airtel is currently doing? I just don't get it. Why not light up more and introduce some sane broadband tariff plans? Why is India still stuck in the stone age when it comes to broadband? IMO, new new 512 Kbps broadband definition ain't gonna change anything.
With Dhirubhai ka sapna Har pocket main Mobile apna, quality of mobile decreased drastically , same will happen if done with internet
( I Know quality still sux, but it can be worsed )
With Dhirubhai ka sapna Har pocket main Mobile apna, quality of mobile decreased drastically , same will happen if done with internet
( I Know quality still sux, but it can be worsed )mgcarley
Founder, Hayai Broadband
IMO, all Indian ISPs are EVIL. What do you have to say about this?
Me? I can't disagree too much. I was greatly saddened when I figured out the situation for myself, having come from broadband paradise.
I still don't understand.
...understand what?
When you have the 100% capacity, why just make use of 1% and in turn impose BS Unfair Usage Policies like what Airtel is currently doing? I just don't get it.
Because 1. Consumers don't make them much money. 2. Capacity theoretically costs money: why "give" to their consumers when they could sell to leased-line customers and other ISPs... why sell twice as much at the same price (leaving less available for later) when you can sell half as much and make more money when the customer wants to upgrade... even if you give him a discount?
But the silly thing is, SMW4 is about 5 years old now - half way through it's expected ROI life, but I suspect that whatever it cost to build ($500 million) has already been made back despite it's actual utilization rates of well below 50% - going by my previous calculations, they only need to charge about $5.30/mbit/month but they charge 20x that much for smaller buyers (up to 622mbit/s I think) or about 12x that much for the high capacity buyers (10Gbit/s+).
Not to mention that the cables in question usually serve 10+ other countries as well, not just India. India would be one of the largest single consumers out of most of the countries these cables serve, though.
Both Tata and Bharti have got their own India-Singapore cables (i2i and TIC) which are operating at ~8 and ~5Tbit/s respectively, but again, utilization must be very low, and I still maintain that if they dropped the costs on this route, they'd still realize similar income - IMO, they should simply multiply the capacities that people are contracted to by 5, and with all this surplus bandwidth ISPs could start offering full-speed ADSL2+ services for no additional cost - or increase the speeds/fups of "unlimited" plans and make a few consumers happy.
If only we were allowed to operate an international gateway... (but we're not, because I'm a foreigner... regulations pretty much stop that proposal in it's tracks).
Why not light up more and introduce some sane broadband tariff plans? Why is India still stuck in the stone age when it comes to broadband? IMO, new new 512 Kbps broadband definition ain't gonna change anything.
That's what I've been saying.
With Dhirubhai ka sapna Har pocket main Mobile apna, quality of mobile decreased drastically , same will happen if done with internet ( I Know quality still sux, but it can be worsed )
I think what they have forgotten to do is increase the capacity on the mobile networks to compensate for the amount of mobile traffic.
As for Internet, with the same amount of users, or even double the users who are paying the same price for better value of service (whether that comes as faster speed or fairer fup or both), I don't think quality will go down so much - so long as the equipment out in the field and the cables themselves are in sufficient condition.
Equipment has design loads sufficient for full-speed traffic between the user and the device (bits per second and/or packets per second) but a fixed amount of physical capacity - you can't plug in 7200 connections in to a device that only has 3600 ports... although upon reflection, I wouldn't be surprised if they've figured out a way to do just that. The uplink on a DSLAM can be anything up to 10Gbit/s
Me? I can't disagree too much. I was greatly saddened when I figured out the situation for myself, having come from broadband paradise.
I still don't understand.
...understand what?
When you have the 100% capacity, why just make use of 1% and in turn impose BS Unfair Usage Policies like what Airtel is currently doing? I just don't get it.
Because 1. Consumers don't make them much money. 2. Capacity theoretically costs money: why "give" to their consumers when they could sell to leased-line customers and other ISPs... why sell twice as much at the same price (leaving less available for later) when you can sell half as much and make more money when the customer wants to upgrade... even if you give him a discount?
But the silly thing is, SMW4 is about 5 years old now - half way through it's expected ROI life, but I suspect that whatever it cost to build ($500 million) has already been made back despite it's actual utilization rates of well below 50% - going by my previous calculations, they only need to charge about $5.30/mbit/month but they charge 20x that much for smaller buyers (up to 622mbit/s I think) or about 12x that much for the high capacity buyers (10Gbit/s+).
Not to mention that the cables in question usually serve 10+ other countries as well, not just India. India would be one of the largest single consumers out of most of the countries these cables serve, though.
Both Tata and Bharti have got their own India-Singapore cables (i2i and TIC) which are operating at ~8 and ~5Tbit/s respectively, but again, utilization must be very low, and I still maintain that if they dropped the costs on this route, they'd still realize similar income - IMO, they should simply multiply the capacities that people are contracted to by 5, and with all this surplus bandwidth ISPs could start offering full-speed ADSL2+ services for no additional cost - or increase the speeds/fups of "unlimited" plans and make a few consumers happy.
If only we were allowed to operate an international gateway... (but we're not, because I'm a foreigner... regulations pretty much stop that proposal in it's tracks).
Why not light up more and introduce some sane broadband tariff plans? Why is India still stuck in the stone age when it comes to broadband? IMO, new new 512 Kbps broadband definition ain't gonna change anything.
That's what I've been saying.
With Dhirubhai ka sapna Har pocket main Mobile apna, quality of mobile decreased drastically , same will happen if done with internet
( I Know quality still sux, but it can be worsed )I think what they have forgotten to do is increase the capacity on the mobile networks to compensate for the amount of mobile traffic.
As for Internet, with the same amount of users, or even double the users who are paying the same price for better value of service (whether that comes as faster speed or fairer fup or both), I don't think quality will go down so much - so long as the equipment out in the field and the cables themselves are in sufficient condition.
Equipment has design loads sufficient for full-speed traffic between the user and the device (bits per second and/or packets per second) but a fixed amount of physical capacity - you can't plug in 7200 connections in to a device that only has 3600 ports... although upon reflection, I wouldn't be surprised if they've figured out a way to do just that. The uplink on a DSLAM can be anything up to 10Gbit/s
- Messages
- 493
- Location
- NA
Just little bit out of discussion, doesn't @mgcarley get pain in hand writing long replies BTW I always wait for his replies, they are really informative 
BTW I always wait for his replies, they are really informative mgcarley
Founder, Hayai Broadband
As of now i dont think we have any speed based plans ,we have size or in other words GB based plans... Could be in near future we should also be having speed based plans...
By speed-based plans I mean the various incarnations of "unlimited" (with or without FUP) - 256k, 512k, 1m, 2m - if bandwidth were quadrupled, there wouldn't be many reasons** that providers couldn't start offering up to 8mbit/s as the de-facto speed (instead of 2 in almost all cases in India) for the same price: 256k users were magically upgraded to 1mbit/s, 512k users to 2mbit/s and so forth... and for the data-based plans, users with 1GB quotas could now get 4GB for the same price, and it would be delivered at up to 8mbit/s or something like that.
**A few infrastructure changes perhaps: domestic transit rates would also have to change - no good having 10Gbits from Singapore to Chennai and only 2.5Gbit/s to whichever city the ISP is in, and the port speeds going to rural areas would have to be increased accordingly as well, but this is mostly a software thing anyway, so no biggie.
---------- Post added at 07:22 PM ---------- Previous post was at 07:17 PM ----------
Just little bit out of discussion, doesn't @mgcarley get pain in hand writing long replies BTW I always wait for his replies, they are really informative
Pain in hand no, pain in back and neck (and a numb arse) from sitting for endless hours in front of my computer, yeah, sometimes. That's why my twitter followers occasionally see messages like this one: Twitter
By speed-based plans I mean the various incarnations of "unlimited" (with or without FUP) - 256k, 512k, 1m, 2m - if bandwidth were quadrupled, there wouldn't be many reasons** that providers couldn't start offering up to 8mbit/s as the de-facto speed (instead of 2 in almost all cases in India) for the same price: 256k users were magically upgraded to 1mbit/s, 512k users to 2mbit/s and so forth... and for the data-based plans, users with 1GB quotas could now get 4GB for the same price, and it would be delivered at up to 8mbit/s or something like that.
**A few infrastructure changes perhaps: domestic transit rates would also have to change - no good having 10Gbits from Singapore to Chennai and only 2.5Gbit/s to whichever city the ISP is in, and the port speeds going to rural areas would have to be increased accordingly as well, but this is mostly a software thing anyway, so no biggie.
---------- Post added at 07:22 PM ---------- Previous post was at 07:17 PM ----------
Just little bit out of discussion, doesn't @mgcarley get pain in hand writing long replies
BTW I always wait for his replies, they are really informative Pain in hand no, pain in back and neck (and a numb arse) from sitting for endless hours in front of my computer, yeah, sometimes. That's why my twitter followers occasionally see messages like this one: Twitter