Internet Protocol Television

Video Overview and Broadcast Video Architecture

James Farmer , ... Weyl Wang , in FTTx Networks, 2017

Internet Protocol Television

Internet Protocol Television (IPTV) is favored by those who don't have experience with broadcast technology but who do have experience with data networks. We must be very quick to point out that a data network is not necessarily a suitable delivery network for IPTV due to the inability of IPTV program streams to tolerate the variable delay of a data network. This issue is covered in much more detail in the next chapter, but we repeat it here because of the number of people we have seen get in trouble because they didn't understand video's needs.

The other issue concerning IPTV is one we mentioned above, and that is the fact that you are going to have to get data to each TV location in the house, and very few houses have suitable wiring. So you are either going to have to provide a category 5 or 6 cable to each TV location, or you are going to have to bear the expense and headache of installing a data-over-coax solution such as MoCA i or HPNA. ii Both technologies convert Ethernet to RF, and form a network using existing coax wiring. Either can be made to work, but either can have issues with using the coax wiring in a way it was not intended to be used—every home is different; some homes will install in a breeze, others will not. Finally, some people think of using the more advanced forms of WiFi to connect TVs. This may work in some cases, but our experience is that putting video over WiFi is going to lead to problems in many (though not all) installations, not to mention security issues.

Think hard about multidwelling units (MDUs) in your service area. They will require rewiring for IPTV (and data), and will you be able to absorb the cost of rewiring? Will you have access to the units, or will the owner deny you access to do rewiring? A good consultant, who has expertise with all aspects of dealing with MDU issues may be an excellent investment.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780124201378000111

IPTV Architecture

James Farmer , ... Weyl Wang , in FTTx Networks, 2017

So in Summary, How Do We Optimize It?

The first issue with IPTV is to make sure you have enough capacity to make all of your customers happy. Video can suffer some delay, but it is not very forgiving of packet arrival jitter. If you put longer buffers in your set tops (usually not under your control), this can make up for more jitter, but at the expense of slower channel tuning, and this makes subscribers angry. The problem is that this is a hard number to predict: you have a mix of unicast and multicast, you have pictures going to big-screen TVs that take lots of bandwidth, and you have pictures going to phones with very small screens, and correspondingly less need for data.

Often the biggest usage time will be during station breaks, and commercial breaks in highly watched sports events. At those times, subscribers are going to start channel surfing, and that will place quite a load on your system. The best guideline we can give you is that you need more than 1   Gb/s just for video, in order to serve an OLT with a couple of thousand subscribers. Ten Gb/s works well, and redundant connections are important. One Gb/s on a PON with up to 64 subscribers seems to work fine. We strongly recommend that you monitor your capacity utilization at every interface possible, and be prepared to add facilities as you see usage creep up. You should consider a bandwidth increase as soon as peak utilization reaches 80% in order to prevent congestion.

For the downstream streaming signal, you must give it the correct priority, usually below control packets and voice priority but above everything else. The priority must be added at the entrance to your network, not part way through; don't let nonprioritized video go through your router and put prioritization on at the OLT—the OLT may do it, but you can get into trouble in your router. Strip the prioritization off at the ONT at the home, and not before. Every device that handles the IPTV must honor prioritization (or another QoS mechanism if you are using something else).

Probably the biggest mistake we have seen systems make when it comes to IPTV goes back to the old untruism, "a bit is a bit is a bit." We say it is an untruism because, well, despite what folks said in the beginning, it is not true. This manifests itself in a manager assuming that just because he has a great IT guy, that the guy is qualified to configure IPTV. "It ain't necessarily so." IPTV is enough of a different animal that you need someone who knows it to configure it. Larger systems may have an IPTV engineer on staff. Smaller systems may send their IT person to their FTTH vendor's training to learn how to configure IPTV, or they may rely on an outside consultant. But if you rely on the outside help, make sure that the help is qualified. We have seen consultants who claimed to understand IPTV do quite a number on unsuspecting systems, things that ultimately cost the system a lot of money; they paid once to have it done wrong, paid a second time to have the wrong undone, and paid again to have it done right by someone who knew what he was doing. Your FTTH vendor or your video system vendor may be able to train your person, or may have someone who can do the required configuration. It'll be real expensive to save the money and not hire them to do their thing for you.

OTT video presents an interesting conundrum. Your subscribers are likely to expect the same performance from OTT as they do from your video service. Yet there is no concept of prioritization or of multicast on the Internet, so you have no control over what happens to packets before you get them. You can and probably should do deep packet inspection in order to determine what packets contain OTT video, and assign them to the same priority as to your video. But the results are not likely to be as good because of what happens before you get the packets. Explaining this to customers or regulators may be difficult. As this is written, the FCC has said that they intend to regulate IPTV and OTT as a data service, which probably means some expectation of quality. But how this will really play out is still up in the air. The saving grace may be that this book may be obsolete and you may be retired before all of the expected court challenges run their course <grin>.

If you are a larger system, then some popular OTT vendors may be interested in setting up servers in your system which capture their more popular content, then stream it to your viewers from the local server. This improves the experience for the OTT viewer, and it reduces the amount of data that you have to get from the Internet, so it seems like a win–win situation. However, keep your ear to the ground regarding regulations coming from the FCC or elsewhere in the government—we are not sure if this practice will be found to be legal or not.

Regarding scrambling (meaning, rendering the video and audio unintelligible without having the descrambling key), it is true that FTTH systems can and do scramble data sent to subscribers. Some operators like to use this scrambling to control who gets what video, and to save the cost of video scramblers in the headend and descramblers in the home. We do not recommend this approach. While you might make it work, the scrambling done in the set-top box market works seamlessly with billing systems, and can respond to high loads of people tuning to an event at the same time. It is designed to work well with both unicast and multicast. The same cannot be said for the scrambling of the PON system. Its purpose is to give comfort to each subscriber that his or her data cannot be seen by anyone else. Let the FTTH system do what it does well, and let the set-top box system do what it does well.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780124201378000123

Application

Vinod Joseph , Srinivas Mulugu , in Deploying Next Generation Multicast-enabled Applications, 2011

Publisher Summary

IPTV is an application that involves multiple technologies. The acronym is defined as "Internet Protocol Television," which would mean that any form of video carried over Internet Protocol, either part of the way or the entire distance, would qualify as IPTV. The focus of this chapter is to discuss IPTV as an application in a multicast environment. IPTV architecture can be broken down into four domains: content domain, head-end domain, transport network domain, and home network domain. IPTV architecture for each major operator is a walled garden, with a customer-engineered solution from the head end to the end user. ITU-T, DSL Forum, IETF, and ETSI are among the organizations contributing actively to the IPTV evolution. A combination of Carrier Ethernet in the core and aggregation and DSL in the access is common transport architecture for IPTV delivery. SSM and IGMPv3 are protocols that together provide the necessary support for users to Join and Leave multicast groups. Quality of service and quality of experience are important parameters in the area of IPTV. Unlike in the data world, the end-user sensitivity to video is quite high so it is necessary for the network to be optimized for video delivery. This chapter discusses all these topics in varying degrees of detail.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780123849236000065

A context-aware system for efficient peer-to-peer content provision

E. Markakis , ... A. Bourdena , in Pervasive Computing, 2016

Acronyms

CATI

content-aware transport information

CDN

content distribution network

CP

content provider

DASH

dynamic adaptive streaming over HTTP

EU

end-user

EUT

end-user iteminal

HB

home-box

HTTP

hypertext transfer protocol

IP

Internet protocol

IPTV

Internet protocol television

JSON

JavaScript object notation

MPEG

moving picture experts group

NP

network provider

P2P

peer-to-peer

QoE

quality of experience

QoS

quality of service

RTP

real-time transport protocol

RTSP

real-time streaming protocol

SDP

session description protocol

SIP

session initiation protocol

SP

service provider

SR

service registry

STB

set-top box

SVC

scalable video coding

TCP

transmission control protocol

UGC

user-generated content

UP

user profile

URL

uniform resource locator

VoD

video on demand

WS

web service

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B978012803663100005X

Class-of-Service Requirements for Quad-Play Networks

Vinod Joseph , Brett Chapman , in Deploying QoS for Cisco IP and Next Generation Networks, 2009

3.11.4 Configurations and Recommendations

Figure 3.14 illustrates a configuration for the Broadcast Video class. In this illustration, the traffic is assigned to a strict priority/LLQ, since the platform in this case supports dual-priority queues.

Figure 3.14. Broadcast Video Traffic in LLQ

Table 3.3 illustrates the recommended characteristics for broadcast video traffic.

Table 3.3. Recommendations for Broadcast Video Traffic

PHB Marking WRED Queue Limit Queue Bandwidth
Expedited Forwarding or Assured Forwarding PHB CS5 as per the Cisco modified RFC-4594 model No; packet drop for this class is not desired Large enough to accommodate maximum traffic bursts LLQ in the case of platforms supporting dual-PQ, or a dedicated nonpriority queue In the case of LLQ, voice and video together should not exceed 33%; if used in a dedicated queue, adequate bandwidth with no "Oversubscription" should be allotted

In Figure 3.15, the traffic within the Broadcast Video class is assigned to a nonstrict-priority queue with adequate guarantees. Here we see that the queue has adequate bandwidth guarantees and does not have any WRED (early packet drop), which can impact IPTV and other types of broadcast video applications.

Figure 3.15. Broadcast Video in Non-LLQ

NOTE 3.4

Broadcast and Real-Time Interactive Video

It is very common for SPs to combine both real-time interactive and broadcast video into a single VIDEO queue. In fact, this can be considered a recommended practice. Video on Demand, discussed in the following section, is also a potential candidate for this VIDEO queue, which helps in the consolidation of video applications with near to similar QoS requirements.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780123744616000033

Multicast Routing

Deep Medhi , Karthik Ramasamy , in Network Routing (Second Edition), 2018

8.10 Internet Protocol Television (IPTV) Multicasting

Despite the bandwidth savings with multicasting, it has not been deployed on an Internet-wide scale in recent years. This is partly due to reluctance by Internet service providers to maintain multicast states in routers. MBone was the last experimental deployment of multicasting in the Internet backbone in the early 1990s. More recently, multicasting has seen a recent surge in managed IP networks. One such case is multicasting for Internet Protocol Television (IPTV).

You may note that every television channel is available for thousands of subscribers in the TV provider's distribution network. With the popularity of IP, such TV distribution networks have moved to IP-based deployment. However, the provider maintains full control over such a network; thus, this is an important example of a managed IP network.

As you can imagine, if a good resolution TV channel requires a 4 Mbps bandwidth, then a unicast based approach would require 4 Gbps of bandwidth if a thousand subscribers are simultaneously watching this channel. On the other hand, by taking a multicasting approach, the bandwidth can be contained to simply requiring 4 Mbps for distribution per channel in the entire managed network. Thus, a broadcast IPTV service can be accomplished using IP multicasting with significant reduction in bandwidth required.

For IPTV multicasting, sources are encoders that originates the video stream. However, an intermediate transcoder can also serve as a "transit" source for distribution. Because of the TV distribution service, source specific multicasting (SSM) is common to use in IPTV multicasting, building a core based treed for a source. Along with SSM for multicasting among routers, IGMPv3 for IPv4 or MLDv2 for IPv6 must be used by end receiver devices at the subscribers' homes. For a specific source s, the subscribing nodes need to subscribe to multicast group 〈 s , M 〉 through IGPMv3 〈 s , M 〉 membership. If subscribers in a subnet do not wish to follow source s, then this part of the tree can be pruned. Also, for a large distribution network, PIM-SM is the best choice.

Note that for broadcast IPTV services, IP multicasting is not the only choice for deployment; multicast virtual private networks (MVPN) using multiprotocol label switching (MPLS) is another possible choice. Also, certainly, you can use unicast to provide the IPTV service although this results in much more bandwidth required overall as illustrated earlier.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780128007372000107

Services Delivered by IMS

Arun Handa , in System Engineering For IMS Networks, 2009

7.4.1 IMS and IPTV

Internet Protocol Television (IPTV) broadly refers to the system of delivering television content with the IP protocol, instead of the conventional broadcast and cable medium. IPTV aims to provide a secure and reliable delivery of video services. These include live TV, Video-on-Demand (VOD), and recorded TV. These video services utilize an access transport agnostic packet-switched IP network to transport audio, video, and control signals. IPTV is architected over a controlled network to ensure a high level of Quality of Service (QoS) and Quality of Experience (QoE). In other words, IPTV is not just video over specific technologies such as DSL or video streaming over the Internet.

The telecom service providers have found IPTV of competitive advantage in the video space. It has been offered as a bundled service referred to as a "Triple-Play." The transport agnostic nature of IPTV allows it to be delivered over a broadband connection such as DSL, or it can be delivered to a mobile device via GPRS or Evolution-Data Optmized (EV-DO). So where does IMS fit in?

Both IPTV and IMS share a common thread. They started with existing best-of-breed technologies and services. Then they laid these out on an IP network, which is agnostic of access transports. When you compare on surface there is nothing radically different from the way things are already done. What is different is that this now provides a flexible and more adaptable framework, which can easily be merged to provide innovative combinations. This is a key foundation of building a converged service.

While the current service models have focused on the technology-focused services, IMS and IPTV shift this to more personalized services. The current telecommunication service paradigm also steers from the bundling of services to blend them instead and provide seamless composite services. It becomes intuitive to see why IPTV differentiates over bundled services of Triple Play and promises services beyond Cable TV VOD.

The consumer is already used to live TV, channel services, subscribed channels, Pay-per-view, and VOD. In addition, the schemes for timeshifting through DVRs and TiVO and placeshifting such as using the slingbox, where a broadband signal can be retransmitted to a different location via IP, are well used. What the consumers have yet to see are more combinational services that blend in communication services with TV programming. Such services are:

•

Interactive TV and SMS Televoting

•

Web browsing

•

Notifications

•

Video calling

IMS complements IPTV to provide a communication platform and an access to a converged network. The IP fabric and transport independence allow the IPTV infrastructure to appear as a UE. With this service identity established, the IPTV can now get an access to converged networks that IMS enables. New IPTV applications for service invocation, delivery, and management are still required. That is still exciting as it opens new avenues that have been limited with the current generation of wireless and wireline networks.

Let's examine some possible services that can be enabled with an IPTV-IMS combination (Figure 7.24).

Figure 7.24. Blended IMS IPTV applications.
•

Interactive Programming The popularity of American Idol, America's Got Talent, and similar shows around the world has primarily been because of the viewer's vote. For now, the viewer has to either SMS or call with a phone device. Calling at these congested hours results in waiting. An IPTV-IMS combination first eliminates the need for a separate device. The IPTV viewer is registered as a UE endpoint. The viewer can now use the remote to send the text message to vote. This brings a new level of participation to the TV viewer. Further, the IMS-based application server can load balance between SMS or SIP message interfaces, to ensure the peak loads do not cause congestion.

•

Interactive Shopping For commercial programming such as the Home Shopping Network or QVC, the viewer can use the TV as a calling device instead of a phone line. The same applies to local advertising for ordering pizza. This provides more convenience and ease of use.

•

Info Streaming During a commercial break, Mary wants to check if her son Chris has left for home after his basketball practice. She sends an invocation for the locate service from her remote. An IMS application server obtains the location coordinates of Chris, plots it on a map, and streams this as Picture-in-Picture on the TV screen.

The standardization of IMS and IPTV is in the initial stages at the time of writing. The ETS-TISPAN TS 182.027 defines the protocols and interaction for IPTV support by IMS. We examine a generic model based on this foundation, as depicted in Figure 7.25. Since both IPTV and IMS are access network agnostic, the model can work with the supporting IP-CAN for the transport and bandwidth control.

Figure 7.25. The IPTV-IMS model.

The IPTV framework comprises a set of elements to support a set of three basic services:

•

Broadcast (BC), which is the support for live TV services

•

VOD or Content on Demand (COD), which allows the capability to access and view content on demand by the user

•

Network Personal Video Recorder (N-PVR), which allows for storing content and viewing at a later point in time

Each of these services can be logically partitioned into three functions: Service Control, Media Control, and Media Delivery. The Service Control Functions interface to the IMS network at the application services layer. This function provides the IPTV service selection and authorization including credit control. It is responsible for selecting the right IPTV media functions. The IPTV Media Control Function provides the control of the media flows and the media elements responsible for processing them. The media delivery function is responsible for the storage, processing, and delivery of the media flow.

The European Telecommunications Standards Institute (ETSI) model identifies specialized elements for the service discovery and selection. It also provides the necessary functions to access the IMS user profile and IPTV service profile.

While the interaction of these elements provides the delivery of IPTV through the IMS as a platform, how do the personalized services happen? Given the architecture where the IPTV service control appears as an application server to the IMS core network, creating the interaction between multiple application servers is required. The SCIM provides the service interaction between a communication-based AS and the IPTV service control element.

As an example, let's examine how an incoming call indication can be provided during a VoD session. We follow on from a video stream that is currently in progress. When an INVITE for a session is received, the S-CSCF forwards it to the Service Capability Interaction Manager (SCIM). The SCIM determines that this requires service interaction. It checks the service profile to determine whether the called party has selected the option to receive an incoming call indication during the VoD, and is not set to a do not disturb. It then indicates to the IPTV service control server to halt the VoD session momentarily. It forwards the INVITE to the voice application server. Lastly, the SCIM then coordinates with the MRFC to stream the image for the caller to the user.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780750683883000071

Basics of Worldwide Broadband Wireless Access Independent of Terrestrial Limitations

Arun K. Majumdar , in Optical Wireless Communications for Broadband Global Internet Connectivity, 2019

2.5 Summary

This chapter introduces the fundamentals of engineering aspects of free-space laser/optical communications protocols and describes technology concepts to establish worldwide broadband access to the Internet or other networks independent of terrestrial infrastructure. Advanced FSO development are the main drivers for satisfying recent high demands in bandwidth thanks to the recent growth of Internet usage, IPTV, VoIP, and You Tube/Twitter and the immediate future and coming applications in IoT. With the rapid growth of data-centric devices and the general deployment of broadband access networks, high data rates from 10   Gbps to more spectrally efficient 40   Gbps or 100   Gbps/channel are feasible. Some basics of network topology, network type, and components with their functionality as well as examples of networks types and categories are also discussed in this chapter. How the optical networks eventually connected to establish interaccess and connectivity are explained. The next chapter will discuss the last-mile problem (or last-mile bottleneck), describing the issues and potential solutions for connecting high bandwidth from the fiber optic backbone to all businesses with high bandwidth networks.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780128133651000023

Deciding to Build an FTTx Network

James Farmer , ... Weyl Wang , in FTTx Networks, 2017

Organizational Considerations

We have previously alluded to doing an objective assessment of the organizational skills you have versus those you will need. We have seen a number of new players struggle because they misunderstood the match between the skills they had and those they needed. Since every situation is different, we cannot give you a list of skills you lack. Rather, we can maybe point out some issues we have seen in over 12   years of helping subscribers roll out FTTH systems, and a lot of years prior to that helping people roll out other technologies.

You may have the best engineer in the world for managing your SCADA network if you are doing that now. But does he know how to configure your entire data network so that it will pass voice with low latency and jitter? Can he evaluate your network for integrity of IPTV, during times when you have lots of channel surfing because the big game half your subscribers are watching just went to commercial? Remember that "a bit is not a bit is not a bit."

If you are the new entrant to the triple play service, you will be expected to meet a higher bar for good service than do the entrenched players. Understanding the economic difficulty faced by small entities, we strongly urge you to have a lab set-up identical to your deployed ("production") network. If this is impossible due to economics, at least you need an isolated PON in your lab that you can equip whenever you need to deploy new hardware or software. There are so many ways to configure networks, and so many ways that disparate pieces of equipment can interact, that it is imperative to test a new configuration before deploying it to subscribers.

If you have not served residential consumers previously, you will be surprised by what you find in homes. You will need installers who have good people skills as well as who understand all aspects of the services you are offering. Some operators do initial configuration of a consumer's data equipment as part of their service, others offer it as a revenue center for extra charge. You will want to think about how you will service your subscribers' equipment, and what the revenue model for that will be. And be sure to factor in customer service people.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780124201378000020

On optimal media/video distribution in closed P2P-based IPTV networks

Hao Cui , ... Weijia Shang , in Computer Networks, 2014

2 Related work

During the past decade, streaming IPTV media/video data over the Internet has been an area of intense study (see for example [3–5]). A variety of video coding standards have been developed to support IPTV application, such as MPEG-2 [6] and H.264/SVC. The MPEG-2 encodes video stream in a pre-encoded non-scalable (single layer) bit-stream. The scalable video coding (SVC) extension in H.264 encodes a high-quality scalable video bit-stream that contains multiple layers. Therefore, IPTV media object is not only coded as non-scalable media stream, but also scalable media stream.

At the same time, with the increasing volume of IPTV distribution through the Internet, how to build and maintain efficient P2P overlay architecture for media stream has become a hot research topic. Based on the underlying overlay construction, two basic solutions to P2P systems have been devised for IPTV distribution [7]:

The mesh-based scheme [8] uses a pull-based architecture, as shown in Fig. 2. Peers swarm media streams over a randomly connected mesh networks. Each peer pulls the desired media stream from one of the other peers that has uploaded it. At the same time, this peer also supplies its available data to other peers. The P2P systems, such as Narada [9], CoolStreaming [10], PRIME [11], GridMedia [12], PPStream [13], PPLive [14], UUSee [15] and SopCast [16], use the mesh-based scheme.

Fig. 2. Mesh-based P2P scheme.

As shown in Fig. 3, the tree-based scheme uses a push-based architecture – media stream is delivered over multiple tree-shaped overlays. The packets of the media stream are forwarded from the parent node to its child nodes. The P2P systems, such as NICE [17] and ZIGZAG [18], use the tree-based scheme.

Fig. 3. Tree-based P2P scheme.

When the mesh-based and tree-based are compared for live P2P streaming approaches [19], the mesh-based approach consistently exhibits a superior performance over the tree-based approach. The time delay in video transmission not only impacts the Quality-of-Service (QoS), but also affects the Quality-of-Experience (QoE). Therefore, it is very important to guarantee the completion of P2P system media distribution within the playback duration time. In [20], a measurement technique for monitoring the video playback quality in the mesh-based streaming system was proposed. Rather than the P2P architecture or quality measurement technique, our work focuses on how to optimally allocate media segments to peers to minimize the system media completion time of non-scalable and scalable media distribution in a mesh-based P2P network.

There are several P2P systems devised for IPTV applications, such as PPLive, PPStream, SopCast, and GridMedia [21], which offer real-time services and have experienced commercial success. They do not disclosure their internal setup. In addition, BitTorrent [22] has been used widely to distribute large files and videos. BitTorrent treats video objects the same as normal files and does not exploit the properties of video bit-streams. BitTorrent employs a tit-for-tat strategy to fight against free riding, resulting in high throughput of the system [23]. However, it is a general file delivery service and divides media objects into equal-sized segments regardless of the peer bandwidths. In CoolStreaming, one of the most popular pull-based systems, the video stream is also divided into blocks with equal sizes to transmit. Unlike BitTorrent and CoolStreaming, we do not simply divide media objects into equal-sized segments for distribution. In our work, the media objects are allocated or divided non-uniformly to minimize system distribution time.

The fluid model [24] proposed by Kumar et al. provides an optimal bound of streaming performance in a fully-connected P2P network, for a video object that is uniformly coded. PROMISE [25] presents CollectCast, a P2P service for media streaming. Based on the working status of peers, it dynamically switches active senders and standby senders to maintain a satisfied collective network performance for media streaming distribution. In this paper, we address peer heterogeneity by adopting non-scalable and scalable coded media/video representations and developing an algorithm to optimally distribute such layered media/video.

Compared to general Internet-based P2P streaming solutions, IPTV setups are usually characterized by closed-network design, where the server has full information of every subscriber, such as its IP address, uplink and downlink bandwidths on the dedicated networks. Every peer runs the same media distribution algorithm that comes from the setup box. In this paper, our research goal is to design a distribution algorithm to minimize distribution delay for media distribution in a closed mesh-based P2P network. This distribution algorithm is adapted to locate the media segment size to peers based on their bandwidth resources and subscription requirement. This paper is the extension of our previous work as noted in [26,27]. Inside this paper, we provide detail discussion on how to create the Optimal Parallel Layer Distribution (OPLD) algorithm for multi-layer media object distribution. Based on the OPLD model, we add in the study of multi-layer peers' departing/joining which impacts the distribution time in the system.

Read full article

URL:

https://www.sciencedirect.com/science/article/pii/S1389128613003861