Linked-list hybrid peer-to-peer system and method for optimizing throughput speed and preventing data starvation

[0001] Embodiments of the present invention generally relate to the field of Video Over Internet Protocol, and more particularly to Internet Protocol television (IPTV).

[0002] Video over Internet Protocol (IP) is a network-based one-way transmission of video content. A video source transmits video as a unicast or multicast transmission. For a unicast transmission, the broadcasting source replicates the video for each targeted client (or viewer). For a multicast transmission, however, the broadcasting source sends the same signal over the network as a single video transmission, but the video is sent to a selected group of clients.

[0003] Over the last few years, there has been interest in live video broadcasting over IP, which is different from other types of communications on the Internet. For example, one primary difference is that live video broadcasting (or streaming) is a one-way communication scheme instead of a two-way communication scheme.

[0004] Normally, client-server network architectures are used for video transmissions such as Internet Protocol television (IPTV) while peer-to-peer (P2P) network architectures are typically used for file-sharing environments where reduced transmission costs are more important than download speed. P2P networks efficiently distribute transmission costs among the nodes by eliminating dedicated server functionality.

[0005] As shown in FIG. 1, a conventional P2P network 100 includes a plurality of nodes 110₁-110_N (N≧2), where each node operates as both a “client” and a “server” to the other clients on the network. More specifically, P2P network 100 features multiple nodes (electronic devices) interconnected to each other over communication paths 120. These nodes operate as either super nodes or leave nodes. A “super node” is generally defined as an electronic device with large resources of CPU power and network bandwidth such as client 110₁. These types of nodes shield “leave nodes,” which are electronic devices having very limited resources such as client 110₂.

[0006] P2P network architectures possess a number of drawbacks. For instance, for conventional P2P networks, data starvation occurs when the ratio of super nodes to leave nodes fall below a minimum level. In a file-sharing environment, large variations in downloading time are commonplace. However, such variations are not suitable when supporting live TV broadcasting where data needs to arrive at its destination within a specific time period or the user will experience a disruption in video playback.

[0007] Another drawback of conventional P2P networks involves the unpredictable delay experienced when a node joins or leaves the P2P network. A node usually joins a P2P network by connecting to a small group (one or two) of other nodes at a time. During this joining period, if this small group of other nodes leave the network, the node will have to restart the joining process again from the beginning. If a very large number of clients are leaving the network one after the other, a new client joining the network may experience a long delay. This situation can be common in live TV broadcasting, especially at the end of a popular show. Setting high priority to joining nodes over existing nodes does not completely solve the problem because it creates yet another problem by slowing down the exiting nodes from switching to a new channel.

[0014] In general, one embodiment of the invention relates to a linked-list hybrid Peer-to-Peer (P2P) network architecture that optimizes throughput at a desired communication speed by grouping nodes with similar bandwidth speeds into separate networking groups. This contradicts conventional P2P architectures where nodes supporting vastly different communication speeds are placed into the same networking group, which causes less-than-optimal throughput due to increased inconsistency of the data transfer rate. By grouping nodes that are capable of supporting similar data rates and excluding nodes supporting data rates less than a threshold data rate from joining that specific networking group, the linked-list hybrid P2P network optimizes the throughput performance of the specific networking group as well as eliminates free-loader problems common in conventional P2P networks.

[0015] According to another aspect of an embodiment of the invention, the linked-list hybrid P2P network architecture guarantees to maintain the maximum throughput of the communication speed by connecting nodes along a one-to-one chain and by monitoring and removing any node having throughput that may have suddenly dropped below a minimum speed threshold. The removed node can re-join the same networking group when its throughput is determined to consistently exceed the minimum speed threshold. The one-to-one architecture reduces the complexity of monitoring the downloading and uploading speed of multiple nodes and further avoids the adverse effects to node speed caused by other nodes transitioning in or out of the networking group.

[0016] In yet another aspect of an embodiment of the invention, the linked-list hybrid P2P network architecture eliminates the delay experienced when a large number of nodes join or leave a network by allowing joining nodes to connect to servers or super nodes while attempting a connect to another node. This connection to the server(s) or super node(s) ceases after a successful connection to other nodes.

[0017] Similarly, a node that loses a connection to other nodes can immediately connect to a video server or super node while attempting a connection to another node. By using servers or super nodes (as in client-server architecture) during a connecting period (JOIN process) and P2P architecture in other period, the linked-list hybrid P2P architecture takes advantage of fast respond times to a large number of peers, joining and leaving a particular channel while maintaining the low cost solution. This hybrid P2P architecture eliminates the need of complex pre-connection or alternative connection of peers since re-connection time is no longer a critical factor.

[0018] In yet another aspect of the present invention, the linked-list hybrid P2P network architecture prevents data starvation and reduces the delay described during a broken communication chain (caused by node removal or leaving the network). The reduction in delay is accomplished by obtaining data directly from one or more servers or super nodes until the nodes can establish communications over the entire sub-network.

[0019] In yet another aspect of the invention, the linked-list hybrid P2P network architecture reduces a channel-switching delay caused by switching of a television channel, by obtaining data directly from servers or super nodes until the node can establish its connection to the P2P network. Conventional P2P network architectures suffer from large channel-switching delays caused by firewall protection between nodes. Nowadays, most nodes will be behind one or more routers equipped with Network Address translation (NAT) firewall.

[0020] In the following description, certain terminology is used to describe various features of one or more embodiments of the invention. For instance, “node” is generally defined as any electronic device that is capable of receiving, processing and/or transmitting audio/video (AV) content, such as video and/or audio streams for example. Examples of a “node” include, but are not limited or restricted to any computer (e.g., laptop, tablet, handheld, desktop, server, mainframe, etc.), networking equipment (e.g., router, bridge, brouter, etc.), a cellular telephone, a video game player, and the like.

[0021] “Software” is generally defined as one or more instructions that, when executed, cause the node to perform a particular function or functions. These instructions may be stored within machine-readable medium such as firmware or software. “Machine-readable medium” is any medium that can at least temporarily store information including, but not limited or restricted to the following: an electronic circuit, a semiconductor memory device (non-volatile or volatile), a hard disk drive, and various types of portable storage devices (e.g., flash drive, compact disc “CD”, digital versatile disk “DVD”, etc.).

[0022] In general, a “connection” is an established communication pathway between two electronic devices. A “network removal event” is an action that causes removal of a node from the linked-list. As an example, if the node enters into an S5 state (shut-down), this action constitutes a network removal event. Other types of network removal events include, but are not limited or restricted to the following: system crash, changing a viewing channel by the user, an inoperable transceiver or the like.

[0023] Referring to FIG. 2, an exemplary embodiment of a hybrid P2P network 200 is shown. Network 200 comprises one or more linked-list hybrid P2P sub-networks 210 that are structured to combine nodes with similar speed capabilities and within the same general locale. For instance, as an illustrative example, nodes in close proximity to each other and supporting transmission rates exceeding 1 megabit/second (Mb/s) may be assigned to a first sub-network 210 while nodes supporting transmission rates between 1 Mb/s and 500 kilobits/sec (Kb/s) may be assigned to a different sub-network (not shown). Herein, first sub-network 210 includes two or more nodes 220₁-220_N (N≧2) in serial communication with each other.

[0024] As further shown, first node 220₁ receives AV content from a content server 230. Typically, the direct connection between content server 230 and first node 220, is one of a more permanent nature. First node 220, receives and routes the AV content to a second node 220₂. As an optional feature, according to this embodiment of the invention, first node 220₁ may also display or play back the AV content. Second node 220₂ receives the AV content from first node 220₁, and in response, also processes the AV content for display or playback. Since there is yet another node in the linked-link P2P network architecture, second node 220₂ receives and routes the AV content to a third node 220₃. Here, third node 220₃ receives and processes the AV content for display or playback, but is not yet connected to route the received AV content to yet another node along linked-list hybrid P2P sub-network 210.

[0025] Referring to FIG. 3, an exemplary embodiment of a node (e.g., third node 220₃) is shown. Node 220₃ is a computer including one or more output devices (e.g., speakers 300, display 310, etc.) and a main chassis 320 that houses logic configured to process and store data as well as receive and transmit data to other nodes. The transmit/receive logic of node 220₃ includes hardware and software that operate to establish communications with a private or public network. For instance, according to one embodiment of the invention, node 220₃ is adapted for IPTV, namely to establish an Internet connection with a traffic controller server 350 to download AV content (e.g., television programming).

[0026] More specifically, node 220₃ establishes a connection with a web site hosted by web site server 340. The user selects a channel to download AV content, such as television programming for example. Web site server 340 redirects the user to traffic controller server 350. Since server 350 is aware that the requested AV content currently is being downloaded to other nodes forming sub-network 210, namely node 220₁ and 220₂, node 220₃ is added to sub-network 210. As a result, AV content supplied by content server 230 is further routed via nodes 220₁-220₂ to node 220₃.

[0027] Software installed within node 220₃ continues to monitor its throughput speed (input and/or output) to ensure these operations do not fall below a minimum threshold. If so, according to this embodiment of the invention, in response to network removal event, the software would initiate a message to node 220₂ that it is disconnecting from sub-network 210 and would notify the user of the disconnection. Once the throughput speed of node 220₃ returns to normal and is above the minimum threshold, a connection with sub-network 210 would be re-established. Otherwise, after a prolonged period of time, node 220₃ may be re-grouped with another linked-list P2P sub-network supporting a lesser throughput rate.

[0028] Referring now to FIGS. 4A-4F, an exemplary embodiment of the operations conducted by nodes within a particular linked-list sub-network 400 is shown. These operations involve a JOIN process in which node (P4) becomes a member of sub-network 400 and establishes communications with another node (P3) in order to receive streaming data such as IPTV programming.

[0029] As illustrated in FIG. 4A, a first node (node-P3) 410 issues a REQ_CONNECTION message 450 to traffic controller server 420. Node-P3 410 transmits a REQ_CONNECTION message 450 after joining sub-network 400 where this transmission occurs prior to a second node (node-P4) 430 attempting to join sub-network 400. Server 420 does not send a response to REQ_CONNECTION message 450 until node-P4 430 joins sub-network 400.

[0030] As shown in FIG. 4B, according to this embodiment of the invention, node-P4 430 sends a REQ_START message 455 to server 420 to request permission to join sub-network 400. Upon determining that node-P4 430 possesses suitable device characteristics to join sub-network 400, such as bit rate as an example, server 420 returns an ACK_START message 460. ACK_START message 460 includes a unique IP address assigned to node-P4 430.

[0031] Thereafter, as shown in FIG. 4C, node-P4 430 sends a REQ_CONNECTION message 465 to server 420 in order to commence establishing a connection to sub-network 400. In response, server 420 returns an ACK_CONNECTION message 470. ACK_CONNECTION message 470 includes an IP address of node-P3 410, which is used for establishing communications between node-P4 430 and node-P3 410. Moreover, server 420 returns an ACK_CONNECTION message 475 to node-P3 410 with the IP address of node-P4 430.

[0032] After receiving these IP addresses, node-P3 410 and node-P4 430 each issue CONNECT messages 480 and 485 in order to (i) pierce existing firewalls (represented by dashed lines) and (ii) establish communications between these nodes as shown in FIG. 4D.

[0033] Generally concurrent with these operations, as further shown in FIG. 4D, node-P4 430 issues a REQ_DATA message 490 to a content server 440 requesting data to be downloaded while communications are being established with node-P3 410. REQ_DATA message 490 includes the IP address of node-P4 430 and a starting sequence number (SeqNum) of the data stream. SeqNum is used to identify from a particular frame of AV content from which server 440 will begin to transmit the AV content to node-P4 430. Such transmission will continue until node-P4 430 closes the connection as shown in FIG. 4E. The connection is closed upon establishing communications between node-P3 410 and node-P4 430 as shown in FIG. 4F.

[0034] Referring now to FIGS. 5A-5D, an exemplary embodiment of the operations conducted by nodes within a particular linked-list sub-network 500 is shown. These operations involve a RE-JOIN process in which a node (P3) 510 experiences a network removal event and disconnects from sub-network 500 as shown in FIG. 5A. Neighboring nodes (P2, P4) 520 and 530 detect that node-P3 520 has disconnected from sub-network 500 and initiate a RE-JOIN process.

[0035] As illustrated in FIG. 5B, node-P2 520 transmits a REQ_CONNECTION message 550 to traffic server 420 upon detecting that node-P3 510 has disconnected from sub-network. Concurrently, node-P4 530 transmits a REQ_DATA message 555 to content server 440 requesting data to be downloaded while communications are being re-established with sub-network 500. REQ_DATA message 555 includes the IP address of node-P4 530 and a starting sequence number (SeqNum) of the data stream. Also, node-P4 530 transmits a REQ_CONNECTION message 560 to traffic server 420 to re-establish connections with sub-network 500.

[0036] Upon receiving REQ_DATA message 555 from node-P4 530, content server 440 begins transmission of the data stream 565 starting at a frame references by SeqNum. Moreover, traffic server 420 returns ACK_CONNECTION messages 570 and 575 to node-P2 520 and node-P4 530, where messages 570 and 575 include IP addresses of their neighboring nodes for re-establishing communications.

[0037] Referring now to FIG. 5C, after receiving ACK_CONNECTION message 570, node-P2 520 issues a CONNECT messages 580 addressed by the IP address of node-P4 530 in order to establish communications with node-P4 530. Similarly, node-P4 530 issues a CONNECT messages 585 addressed by the IP address of node-P2 520 in order to establish communications with node-P2 520. Once communications are established, node-P4 530 begins to receive AV content from node-P2 520 and terminates receipt of the AV content from content server 440 as shown in FIG. 5D.

[0038] Although the present invention has been described with respect to specific examples and subsystems, it will be apparent to those of ordinary skill in the art that the invention is not limited to these specific examples or subsystems but extends to other embodiments as well.