The Pros and Cons of Open DSP architectures

hello and welcome to dsp leaders world forum 2021. coming up now is our challenges and opportunities round table where we'll be discussing the pros and cons of open dsp architectures we are at a critical point in the evolution of telecoms networks on the one hand it is so encouraging to see service providers getting on board with open telco ideals as well as vendors most of them and policy makers but there's still plenty of doubt and uncertainty around will those aspiring to create truly open digital services providers create the blueprint for the future of our industry or is this all just wishful thinking well let's find out what our special guests think and i'm delighted to introduce franciser who is vp for access to segregation at deutsche telekom dr mike short chief scientific advisor department for international trade with the uk government alex resnick outgoing chair of the etsy mec isg and distinguished technologist at hpe and rick hamilton senior vice president of siena's blue planet software hello everyone and welcome to our state of the industry debate on the open telecom now during this session i'm going to put a few topical statements to you all my colleague ray lemaitre and i drew up these talking points based on the feedback from our telecom tv community so contentious perhaps but certainly relevant and indicative of what our viewers are asking us about so let's hear our first statement then far from being an attractive economic option for telcos an open architecture will lead to greater costs due to increased interoperability testing integration and deployment issues i'm hoping you can all answer this one but but franz let me start with you because i introduced you you first to our audience what do you make of that do you think that such an architecture actually far from being cost effective comes with a lot of additional cost for telcos it comes with additional complexity but additional complexity if you deal with it smartly and we have to uh should not leave lead to additional cost i would even say must not lead to additional costs so it's it's of course if if you if you uh disaggregate and and put things in smaller pieces and then you put them back together again of course you need to take care and make sure things work and as you said there are many apis you need to verify um but nobody said and i think that's not the basic idea that everyone uh every telco who wants to implement that starts from crash scratch and does everything uh itself so of course it's all about community and collaboration um that's why we have all these community activities out there that takes care that things work together uh as flawlessly as possible so you of course you have the the final the final piece and adapting to the to the specific use case um is something we need to take care about but that's something we do uh today as well but of course the expectation is there is a kind of pre-integration already happening as a community effort across all the players um because it's the only way from my perspective that the smaller software only player uh can can play a significant role in the in the rather big overall ecosystem we need to put up our services and provided in the required quality so yes additional complexity but if you deal with it smartly and of course things like automation also come in place here um we strongly believe that in the long term and overall it will definitely lead to lower cost not higher cost okay thank you friends for that i wonder what our other guests think rick maybe i could start with you are we simply replacing one set of costs with another like to like set of costs here i'm not sure i'd view it that way i think uh as front said it's sort of a matter of perspective you know when do you think about cost and how do you think about opportunity cost and the cost of capital so like anything in technology you know one one one sort of architecture is never going to fit anybody but i tend to agree with him that likely the likely problem will be in the short term there'll be some complexities some new challenges to deal with but long term you know having more disaggregation more open sort of architectures provides for greater choice it provides hopefully for easier upgrades and evolution of technology without massive retroactive changes so there's a lot of attractive long-term benefits to to a more open architecture and you know it's difficult to do with technology but i think if we look over the long term it's going to be better better for everybody that's sort of how i view it thanks rick alex so should we be looking at taking a long-term view on this one we shouldn't get too fixated with any uh short-term spikes yeah absolutely um i think you definitely need to take a look at that long-term view the initial transition into a different way of doing things is always costlier at that initial point but at the end of the day so much of the cost is in the operations which are naturally long-term and as we've seen throughout the rest of the technology world a transition to an open approach a transition to an approach that allows you to pick the best of breed to operate things in a way that makes sense to your business always at the end of the day results in both lower costs but also potentially being able to address um new opportunities much faster and much more efficiently right so you also win on the revenue side of of that equation in the long term thanks alex mike what's your perspective here because you you've seen that you've seen a lot of uh technology changes and evolutions here um and this this being the latest one so before i join the government i work with telefonica and its predecessors for 30 years and we've actually seen open architectures since the days of 2g and gsm what we're now seeing is a much greater openness because of softwareization and standards and that's extremely helpful when it comes to both cost reduction but also time to market the ability to deliver new products and services more quickly we have had open architectures in areas like distribution and billing for quite some time so it's good to see it's coming to the network world i think we've also as operators and i was an operator once we've we've ingested ideas as they've come along so uh the wi-fi activity alongside cellular from the days of the smartphone from 2007 onwards so i think the the move towards openness is good i think the move towards open data is great but i think the standards basis is absolutely key to minimize any complexity and to maximize the opportunities for cost reduction again horses for course need the right blend of talent uh for the 21st century thanks mike well thank you all uh let's move on to uh our next point here and that is we've been looking at the issue of security the all-important matter of security now the more vendors you have in the network the more interconnect points there are the more opportunities there are perhaps for security breaches or to put it more simply is an open network more inherently insecure uh so are we increasing the risk of security issues as we go along the open road rick i'm going to come to you first because i think you're going to shake your head vigorously at this one yeah no i don't look i think the the the tidal wave of security is upon us and the number of vendors you have in your network while interesting doesn't in my mind solve the fundamental problem i mean if you think about sort of forget open architectures for a minute if you think about the promise of things like 5g a big push to sort of edge computing and certainly the proliferation of you know applications and devices at the end and edge of the network there's where the real security uh problem exists and i think it requires a holistic and different approach you know think of things like sort of a zero trust posture in your network so i think it's pretty sh i believe it to be short-sighted to say well if you have more vendors through these open architectures you're going to have a bigger security risk it's coming it's in fact it's already here today just with the proliferation of devices and the advancements of our ability to put connectivity way out to the edge we've got to think about this differently so it's an argument that i tend to kind of say wrong wrong question for the wrong answer if you know what i mean that's that's how i view it great thanks alex please so i'll i'll add to to that um i think there's kind of a misnomer here right that or miss miss idea that you go when you go from a single vendor or a few vendors to an open architecture because you have so many different entities bringing pieces together you're now increasing your security your security risk but when the telcos used to buy and still do right very highly integrated appliances those appliances are put together for many many different components um essentially many different vendors sometimes in reality different vendors right but sometimes even if it's just different groups within a single large organization from a security point of view these are um pieces that are being brought together the difference with an open architecture is you have a well-defined well-thought-through interfaces at where those pieces are being put together you also have visibility into what's going on there understanding of what's going on there so you actually have better information to both uh build a more secure system and address um and mitigate security risks as they come up once the system is operational in an open architecture then you would have in a closed architecture where you as the entity that at the end of the day operates the stuff you really have no information and no visibility and no access into some of the key interfaces that uh your attackers might be exploiting right so i think it's actually the opposite way around i think an open um at an open architecture when properly understood and looked at from a security point of view with the right security stance like a zero trust approach to everything gives you a much more secure network thank you alex and we will be coming back to those points about the ecosystem and also um interfaces um a bit later on in our discussion um mike do you see the same thing here do you see that uh actually what this is actually giving us is greater i visibility he's giving us more solutions as well as visibility um and i think if you have more supplies you have more ideas that can help you but i'd like to just apply the lens of the customer and the regulator i think from the customer point of view the customer will expect no different in an open architecture world to today's world if anything he would expect improvements so from a customer viewpoint it's quite important to reassure the customer from a regulatory point of view and operators are regulated we also need to take the regulator with us and the regulator often measures things like security attacks or malware and also the resilience of networks so again the regulator needs to be taken with you on this journey so it brings great opportunities with security new ideas but we need to take certain stakeholders such as customers and regulators with us on this journey thanks mike and franz um is disaggregation and this proliferation of vendors and choice actually assisting with security matters yes yes yes so uh when i talk to my security experts the very first thing they tell me it's not only a wrong but a very dangerous assumption uh believing a closed system is more secure than an open system as already explained in an open system is open interfaces many more people can look at it you you can much easier verify if if things go wrong in a closed system most likely you don't even recognize you may recognize something is behaving strangely but you have no means to to verify and to figure out because you don't understand the system internally so this believing something you don't understand is secure is is i said a very very dangerous assumption uh so the openness is is clearly uh helpful based on couple of key assumptions as i think also already very well explained so you the the apis and interfaces need to be well understood well documented uh and can be tracked easily and then you should be able to isolate the the source of of risk uh much easier going forward always taking uh into account how complex the overall systems are we are operating here great thanks everyone for that um this next point i'm going to address come straight back to you friends on this one because it's very much telco focused and it's it's something we are we hear about a lot on telecom tv our audience is talking to us about this quite a lot that is open architectures promise more diversity in the supply chain but vendors are smart and those who've already got the extensive supply contracts aren't going to give way easily and they may bundle or pre-integrate components or they may buy out the emerging competition we're seeing both of these things happening already so telcos will mostly end up working with exactly the same suppliers as they do now is is this just really another way for toco's to put a bit of pricing pressure and cook costs on their their existing vendors of course cost is always a topic for us that's that's i think um well known but desegregation uh is is much more uh it's uh yes and it's honestly it's good news if we if we see uh the existing uh suppliers are going down that route as well um and they may very well be in an excellent position to to cover uh a broad range of cases but it's always a question where are the uh leading and they are not there if if you go with a big supplier you always have to compromise at some point in time because i said this the system itself is is so complex and so many different components need to come together uh it's it's a bit hard to believe that there is one single company that that can be uh the best in all the aspects and openness gives me exactly the opportunity to work in certain cases with more advanced companies working more tailor-made cases also keep in mind if there are only a few suppliers providing solutions for the whole world so to say um they always need to compromise yeah they cannot they cannot implement all these this nitty gritty special cases that around which are so important for us to differentiate and different markets have have different priorities and then what goes well in one market may not go so well in the other market but if you only have a few suppliers you either convince one of them to to do it or it takes your time and that's exactly this aspect of speed and being able to provide new solutions in much faster time that's that's exactly the sweet spot where you'll start uh working this with smaller suppliers and smaller companies which if you have an open architecture you can you can easily uh plug in your overall system if few of them hopefully manage to become really relevant suppliers uh for the overall ecosystem what i very much hope as we had that situation some time ago as uh some of us for sure remember that would be of course the best case but that's of course something the the market will decide it depends on what products they're delivering however reliable they are how good they react so the typical market mechanics of course will come in on the point of takeover yeah that's that's of course a difficult one if there would be a good recipe to that of course that would be good i mean we we know in many industries uh takeovers are sometimes an issue because they happen because of different motivations where it's hard to to do a lot but of course there's something we hopefully find find some some ways to to [Music] avoid that every promising uh small company is is taken over more or less immediately but difficult thanks very much friends let's go straight to our two two vendors on our round table today and see what they're thinking so far about um these open open turco architectures and models um alex let me come to you to you first um obviously this is a big opportunity um but do you think things will really change do you think opportunities will increase uh yeah i i mean i think this the this this the focus on you know big companies and take over smaller companies is a bit wrong uh i think what you got to focus on is whether you're driving innovation into the market how broad of an event of an innovation and how fast that innovation is being driven and i think clearly in our industry the last let's say decade the speed and the rate of innovation um the way people try to do things very differently has increased tremendously and a lot of it is being driven by open architectures across everything that we do in telco right um and that creates opportunities for small companies now you know we have to remember right small companies if they succeed they become large established vendors so to me i don't know that it's that important whether a successful new entrant becomes large by being absorbed into a legacy player uh but thereby really changing the way that legacy player approaches the market or whether they do it organically and they grow to be a large company right what's important is going back and looking at um you know have we been successful in driving the speed of innovation and i'd say in the last uh if you look at the telco space over the last 10 years the answer is absolutely yes and to a very significant degree that's driven by open architectures across the board thanks alex and rick what do you feel on this one yeah i think alex is right i mean i agree that um for me i think the most important change will come in the fact that as these open architectures get adopted and i think we all believe that they will uh it creates an opportunity for choice and choice drives a competitive behavior and competitive behaviors bring innovation to the market uh it forces the big you know network equipment providers uh even sienna being one of them to really think long and hard about where they're spending money where they can be the best uh for an overall solution what is their real core competency and redirect you know those capital dollars into that innovation i don't think we can lose sight of the fact i guess i'm biased because i've spent maybe the past 20 years of my career working for the big vendors cisco and cnn and others uh they've put or we've put billions and billions of dollars into r d that have really driven important innovations into the market and you know we don't want to lose that for sure but i think the open architectures will as alex said it will give people the ability to make choices based on what part of the architecture are they working on what problems are they trying to solve and who could be the best in the market i think that will only put the big vendors uh on notice they'll have to continue to drive innovation to be competitive and as technology you know given the pace that technology is moving at that competitive innovative uh drive is going to help everybody so i i really see this as a positive for the entire industry it's going to allow the small guys to get in and really disrupt and it's going to force the big players to really uh you know think carefully about where they're spending their their our lead dollars to drive the biggest impact thank you rick yes it's very interesting that you know maybe it does for uh force the the larger vendors to focus on their core competencies as you say mike do you think we'll really get an uptick in in innovation especially from new entrants smaller companies maybe spin-offs from academia i think we're already getting an uptick uh there are more uh companies interested in 5g currently than we've seen in previous generations it's true that there are some significant large vendors in the mix but there are many uh startups that are interested and that's because we're now in a world of a digital economy it's no longer just about networks it's also about applications it's also about cross-sector activity i was looking at some fintech companies the other day uh some insure tech and some health tech companies and and they are using mobile as if it's always been with us um the reality is that open architecture and open standards helps to give us faster routes to market for many industries it also helps to differentiate between perhaps one operator and another should they wish to do that it also needs to be looked at alongside other measures that are being considered so we're seeing a growth and interest in site sharing and network sharing and even some neutral host sites so we need standards that can work in a in a different way to say 20 years ago standards that are more open and more flexible but recognizing that capital the investment required is just not growing on trees so i think overall open standards and open architecture are the only way forward thank you mike and this leads us very nicely into our our next uh statement that we're we've been discussing with our audience um and it refers to something that uh certainly alex mentioned earlier an open architecture relies on apis and maintaining consistent open apis across the whole communications networking ecosystem will simply prove not to be possible this has been really difficult so far various organizations that have are still really pushing pushing this but complete industry support and avoiding fragmentation and different versions of implementations is tricky alex do you think we'll we'll achieve this yes and and so i am completing my my term as chair of etsy mac um that and it's a standards body that defines some of the key apis in the mac space so this is a topic that's very near and dear to my heart and the answer is absolutely yes and we've done it as an industry consistently i mean look apis are interfaces right there aap and then i stands for interface standard standardized interfaces where you need global consistency and interoperability that's their job this industry has been super successful at doing that uh 3gpp right it's a um at the end of the day it's an interoperability interface standards that's his head of standards um 802.11 for wi-fi right global wireless lan standard um ip right it's an interface protocol between two devices right so um the industry has done it successfully where there was a need for global consistency that has always been arrived at and i have no doubt that we're gonna do it for the areas where it's necessary right so so so the only i think the only thing that we'll find out is where the market demands it and where the market does not demand it if the market does not demand that kind of consistency then it's not going to happen if it's not needed all right a very positive response there um let's say uh rick what what's what's your take i mean there will always surely there'll always be um some more bespoke and interesting implementations of interfaces that just just happen yeah it just happens but i tend to agree with alex the industry has proven this out you know many times over the past couple of decades and he listed a number we we actually see you know in our space from blue planet software automation machine learning it is absolutely dependent upon the development of these standards because if they don't start to emerge the hopes that we have around you know just what i would consider to be basic and core automation are going to fail if those things fail the adoption of these advanced technologies like 5g you know what's going to happen with the edge that they're they're going to risk you know an adoption problem so i think it's almost a must right these standards are going to emerge uh and they're going to change the game for everybody i i do think the pace at which they emerge is the open question and it's the age-old problem if um all of the vendors in the world today and some of the service providers get together and try to debate these standards over and over again we'll have an elongated time frame really to find a place where they're having an impact if the csps the dsps can start to come together you know as the consumer of these interfaces and these standards they'll greatly increase the pace at which we can adopt them and you know that's one worry we have you know can the major telcos around the world really come together and agree upon a base set of standards and and i think they will and when they do uh we won't be asking this question anymore we'll be looking at sort of the innovation that will be driven through their use sort of how i see that playing out great well you know let's ask let's ask friends um how how how is it going with deutsche telekom and the other major turcos the tier ones are you getting together to um get a court here and an agreement yeah we have to so i very much agree my my philosophy wherever there is substantial benefit we will have um a unique set or a dominant set of apis i wouldn't go as far as saying a unique but a very dominant one the price to pay of course it takes time agreeing the more company the more players you have the more difficult it gets to to agree um what is it but if it's worth the effort and it's very obvious you have to do it then it happens and i said i mean the best uh and most successful example we have is is definitely 3gpp otherwise we couldn't use our mobile phones and all the networks across the world uh and and we are building on that do we need it everywhere no because of i said it makes it slow so so it of course if if you first need to agree on on an api uh on on on global basis and only then you can build of it you have to wait so usually you you start to innovate without that but if things matter then it it will come and will be put and we do that yeah we we work together as as some of you for sure have recognized we we just uh did uh sign and publish an mou between a couple of major players in europe uh regarding oran with a clear intention if you read the demo you that we will as next step now work on the so-called tech priority document where we exactly start to figure out what are these priorities we need to jointly work on across the operators to make sure that this uh desegregated architecture is is uh has has a fair chance to grow and and all the the players in that ecosystem know exactly what they need to do how they can contribute and the things fit together in the end to minimize the friction in the in the early days of that new technology yeah that certainly was a welcome move recently um from those those four telcos uh mike it's a trade-off isn't it let's do to which of the key apis we we need to um spend time getting right and sorting out and the need also to what the increasing need to move fast and be agile yes it's a prerequisite but i think it should be demand driven by the operators and indeed the the broader community but let me just go back to something very simple gsm and the subsequent generations have been based on standards which deliver economies of scale and shared r d so therefore many users benefit from the lowering lowering of unit costs we wouldn't get 1.3 to 1.5 billion

smartphones being shipped every year if we didn't have standards and we wouldn't have close to 300 million 5g customers at the end of 2020 were it not for standards so actually what i see these apis are is an increase to the range of standards adding more capability and the capability to to all users and operators in the mix we also need to think about this in a data world type way and of course it's no longer just about voice there are many more options in the data world so i think we need to embrace the internet i think we need to embrace fixed line communications and we should be promoting the best use of standards including new apis when they come along and give real benefit thanks mike well a final statement before we wrap up this discussion to discuss um we want to look now beyond the physical network and towards the commercial objectives you know the more a telco goes down the dsp route and looks to digital services income and revenue the more they lose focus on their core strength i.e connectivity and therefore the more they put their futures at risk so it's a contentious one this from our audience um it's really looking at should a telco head more towards digital services provision or should it double down on its basically core connectivity strengths and this ties in with this need for for the open network the open telco the open tech is all about opening up and creating more opportunities um out there especially with dsps so franz d do you feel that the open telco and what we've been talking about naturally leads itself towards digital services provision and is there a risk that we take our eye off our core strengths our connectivity actually i don't see that exactly the opposite so our ambition of course is to use all these principles of open architectures to better monetize our connectivity based products so it's of course our our key product is no discussion is is about providing best possible connectivity uh and and having as an open flexible software architecture on top being able to control your connectivity product via via such an open architecture and providing apis opens completely new capabilities uh how to monetize that so this is not is is not is not diluting us the opposite we are using that to to have uh better opportunities uh to to even strengthen our position in that field thank you franz uh let's go around for our final comments alex um do you feel that all this work we're doing on open telco is actually as fran says helping us to monetize what we've already got monetize our connectivity yeah and and i'll i'll bust telecom to these chops a bit on this one i mean what this question really says is a large business like atelco can't focus on two adjacent areas at the same time one always comes at the expense of the other and i just don't i i i i we we've got plenty of examples where that's not true especially when the areas are so interconnected right what what we're driving towards and what telcos are going to be in the middle of is a highly distributed um very power powerful edge cloud now when you go today and you go into any major cloud provider or google and amazon um a microsoft guide and you start setting things up what is one of the first things that you wind up setting up within the context of their centralized cloud um your virtual um your your your virtual cloud which is essentially the virtual network that interconnects the various virtual pieces that you have in there right so um in a digital world the networking piece is absolutely critical to the extent that you have to create one whenever you start any reasonable cloud project in it in a virtual sense right so as we go to a highly distributed edge network that we all agree is gonna drive this next generation of technologies that connectivity is the core on which everything else gets built right so it's not i don't think anyone is taking their eyes off the connectivity piece alex i can't do two things at once so that's why i'm highly skeptical that telcos can do two things at once and mike it's a very good point this isn't it um we can do two things at once i'm more than capable of this um it's not an either or choice here there's there's connectivity and there's there's richer digital services out there yes exactly the premise of the question i think is wrong i think operators particularly mobile operators are much more than just connectivity even though that is a vital ingredient if we look to some markets we can see they're really strong on security or optimizing the secure communications for key customers we can see in other markets that they're really strong on mobile money transfer in other markets we we can see they're really strong on mobile media now that's not in every country that's true and that's partly about market phasing but i think what we can see with open standards is that new markets are opening up so that's another key reason to embrace open standards and open architecture yes there's some complexity there yes security does need to be addressed but it provides more opportunities than just sitting still thanks and rick final uh comment from you the the open telco naturally therefore lends itself towards new opportunities for telcos yeah certainly and and as as the rest of the panel is you know wisely pointed out the the the underlying nature of how networking and connectivity works is going to evolve as these technological use cases evolved i mean just think of things like you know network slicing in a 5g world just to pick a cliche that everybody talks about if a service provider is not actually living the world of delivering those services beyond the connectivity they're never going to understand fundamentally how that connectivity needs to evolve so i think it's critical to their future to be in this game and as was said earlier finding new ways to monetize that connectivity is going to come through how people are using the technology that hangs off the end of all this connectivity so it's going to be going to be critical for them to you know walk and chew gum at the same time it's it's going to be a matter of survival in my opinion i don't think it's a choice excellent rick thank you very much and with that we have to draw our discussion to a close thank you all very much for participating and sharing your views and opinions today now if you are watching this on day one of our dsp leaders world forum then don't forget to send us any questions or views you have on this subject and we'll try and answer them in our live after show program later today otherwise please take a look at all the sessions from this year's world forum we are making them available for on-demand viewing throughout the week today is focused on the open telco but we are covering four other important topics as well as part of this year's event enhancing the ecosystem cultural transformation 5g and sustainability and dsp revenue streams we have a great lineup of speakers this year across our round tables interviews and live q a programs so please keep watching and do get in touch with us and let us know your views for now though thank you for watching and goodbye you

2021-03-14 12:13

Other news