EP2030129A1

EP2030129A1 - Peer-to-peer communication device, method for forming a peer-to-peer identification indication, and method for operating a peer-to-peer communication network

Info

Publication number: EP2030129A1
Application number: EP07729599A
Authority: EP
Inventors: Jens-Uwe Busser; Christian Greibich; Adam Kiss; Adam Kovacs
Original assignee: Nokia Siemens Networks GmbH and Co KG
Current assignee: Nokia Solutions and Networks GmbH and Co KG
Priority date: 2006-05-29
Filing date: 2007-05-29
Publication date: 2009-03-04
Also published as: WO2007138044A1; DE102006024982A1

Abstract

The invention relates to a peer-to-peer communication device comprising a memory in which a peer-to-peer identification indication of the peer-to-peer communication device is stored, said indication comprising a distinct, non-modifiable part and a modifiable part.

Description

description

Peer-to-peer communication device, method for forming a peer-to-peer identification information and method for operating a peer-to-peer communication network

The invention relates to a peer-to-peer

A communication device, a method for forming a peer-to-peer identification information and a method for operating a peer-to-peer communication network.

In peer-to-peer communication networks, a plurality of network entities, so-called peers, implemented by computer systems are interconnected to a communication network structure without the need for a central entity, such as a centralized server.

Peer-to-peer communication networks thus have no classified ^¬ forensic client-server architecture, but each peer acts both as a client and as a server. An application ^¬ example of peer-to-peer communication networks is a Fi ^¬ le-sharing system in which data is distributed by means of the peers stored and each peer at any other peer can request data stored there.

Since there is no central server in a peer-to-peer communication network, certain tasks that are taken over by a central unit in a client-server architecture must be taken over by the peers themselves. One such service of high importance is the address resolution, ie the implementation of known peer address data (such as the name of the user or operator of the peer, a telephone number assigned to the peer, etc.) required for contacting a particular peer, into the IP ( Internet Protocol) address of the peer. This can be at ^¬ play as vividly be reached by searches, but this is not considered further in the following by "flooding" of the peer-to-peer communication network. Another way to resolve addresses is to use distributed hash tables (DHTs). There are several algorithms for this. The best known are CHORD with an annular logical arrangement of the peers, that is, in the peer-to-peer communication network having a ring topology, and capital demlia, according to which the peers according to an baumarti ^¬ gene structure are logically arranged.

These algorithms are based on assigning each peer a unique address, called the node ID, which defines the (logical) position of the peer in the peer-to-peer communications network.

Data to be stored in the peer-to-peer communication network is identified by specified search terms and stored according to the search terms by means of specific peers. Conversely, then the peer, stored data by means of which, using the data zugeord ^¬ Neten prefixes are determined.

To accomplish this, data associated with a search term is stored using the peer whose node ID most closely matches the hash value of the search term according to a metric used. The procedure for the CHORD algorithm is described in [1]. The Kademlia algorithm is written in [2] be ^¬.

Because the assignment of search terms and thus data to

Peers depends on the node IDs of the peers, the user of a peer can determine over which search terms or data his peer is responsible if you let him freely choose the node ID of his peer. Even though data in a peer-to-peer communication network is typically stored by means of several peers in order to avoid data loss when a peer leaves the peer-to-peer communication network, the node IDs are freely selected by the peer - However, users still risk that a user of several peers (that is to say a user who operates several peers) or that a user group ^obtains complete control over specific data and make these, for example, no longer available.

Therefore, in a peer-to-peer communications network (among other security requirements such as access control to the stored data, authentication of stored records, etc.), it may also be necessary for a user to have a unique user identification (unique user identifier, UUID is assigned) and the peers of the user ^¬ unambiguous network unit identifiers which make it possible to determine which peers are operated by the user.

The UUID of a user is usually derived from a eindeu ^¬ term name of the user, for example, the customer number of the user, his e-mail address, an assigned thereto a SIP URI (Session Initiation Protocol - Uniform Resource Indicator) or its telephone number. The user's name is typically bad From ^¬ line of a UUID as a name may not be unique, as there can be multiple users of peer-to-peer Kommunkationsnetzwerks with the same name. Likewise, the derivation of the UUID from a nickname freely selectable by the user is not recommended since the user is thereby at least subscribed to the peer-to-peer communication network or to the communication service realized by the peer-to-peer communication network ( for example, a file sharing communication service) can even determine its UUID, which may Angriffsmöglichkei ^¬ th of the user to the peer-to-peer communication network created.

To derive the UUID of a user calculates the hash value of one übli ^¬ cherweise (eg according to a Secure hash method, such as SHA-I (Secure Hash Algorithm 1) or MD5) the unique name of the user (for example, his telephone number) and truncates this if ^{necessary ¬} by cutting to the length of the UUIDs used in the peer-to-peer communication network. This UUID can then be used in a digital certificate that allows the user's peer to be peer-to-peer to other peers as an authorized member.

Identify communication network. Thereby, the user's peer can be operated exactly at the location indicated by the node ID specified by the UUID in the peer-to-peer communication network.

Many conventional peer-to-peer communication networks ent ^¬ no access control in the form of a restriction of the choice of the node ID of a peer. In this peer-to-peer

Communication networks, each user can operate a peer with any node ID.

Furthermore, there is the possibility that the entry of a new peer into a peer-to-peer communication network is denied if a peer already exists at this point (i.e., with the same node ID) in the peer-to-peer network.

Communication network is present. Alternatively, a message is sent to the already existing peer that causes it to leave the peer-to-peer communication network and leave the new peer space. However, if the node ID of a user-operated peer matches (or is uniquely specified by) the user's UUID, then it is not possible for a user to have multiple peers in the peer-to-peer communication network simultaneously operates.

In order to make this possible, the user could be certified for each peer operated by his own UUID or his own network unit identification, but this requires a high certification effort. The invention is based on the problem of creating a more flexible and at the same time secure use of peer-to-peer communication networks over the prior art.

The problem is solved by a peer-to-peer

Communication device, a method for forming a peer-to-peer identification information and a method for operating a peer-to-peer communication network with the features solved according to the independent claims.

Exemplary developments of the invention will become apparent from the dependent claims.

A peer-to-peer communication device with a memory is provided in which a peer-to-peer identification of the peer-to-peer

Communication device is stored, which has a unique unchangeable part and a variable part.

Furthermore, a method for forming a peer-to-peer identification for a peer-to-peer

Communication device provided in which the peer-to-peer identification information is formed from a unique unverän ^¬ derbaren part and a variable part.

Furthermore, a method is provided for operating a peer-to-peer communication network with a plurality of peer-to-peer communication devices, in which the peer-to-peer communication devices are addressed in each case by means of a peer-to-peer identification information, the ei ^¬ NEN unambiguous unchangeable part and a variable part has.

Clearly, an embodiment of OF INVENTION ^¬, according dung peer-to-peer communication device specified one for addressing a fixed ID portion extending For example, uniquely results from a unique identification of the user of the communication device. The fixed identification part is transformed into a peer-to-peer by an identification part that can be modified by the user of the peer-to-peer communication device, for example.

Identification information added, so that the peer-to-peer identification information for addressing the identification information communication device can be used in a peer-to-peer communication network.

Under the variable part is to be understood that the user of the communication device may freely choose this part of the peer-to-peer identification, while the ^¬ uniquely unchangeable part of the communication device is permanently assigned and can not be changed by the user. The variable part of peer-to-peer identification data, in an alternative Ausgestal ^¬ processing of the invention also automatically, for example using a corresponding Identifikationsangabe- allocation algorithm are determined and assigned to communication means of the Peer-to-Peer.

The peer-to-peer identification information is thus not fully ^¬ constantly fixed unchangeable, but part of the peer-to-peer identification information is not set or not checked in the Authentifi ^¬ ornamentation or authorization of the communication device.

For example, one embodiment of the invention allows a user to operate multiple peer-to-peer communication devices in a peer-to-peer communication network that can be uniquely assigned to the user, enabling the user to flexibly address peer-to-peer communication devices operated by it. At the same time, the security of the peer-to-peer communication network can be ensured by the unambiguous assignment of peer-to-peer communication devices to users For example, it can be avoided that data is stored only by means of peer-to-peer communication facilities of the same user.

The further embodiments of the invention, which are described in connection with the peer-to-peer communication device, also apply mutatis mutandis to the method for forming a peer-to-peer identification information and the method for operating a peer-to-peer communication network ,

The unique unchangeable part is, for example, a unique user identification of the user of the peer-to-peer communication device. The unique user identification can be determined, for example, by means of a hash function from a unique name of the user. As a unique user identification, for example, a telephone number, an e-mail address, a SIP URI or a customer number of the user can be used.

The variable part is, for example, a freely selectable by the Benut ^¬ zer the peer-to-peer communication device bit combination.

In one embodiment, the peer-to-peer communication device is part of a peer-to-peer

Communication network operated according to the CHORD algorithm or according to the Kademlia algorithm.

The peer-to-peer communication device is, for example, a personal digital assisstant, a desktop computer or a laptop.

Embodiments of the invention are illustrated in the figures Darge ^¬ and will be explained in more detail below.

Figure 1 shows a peer-to-peer communication network according to an embodiment of the invention. Figure 2 shows a flowchart according to an embodiment of the invention.

Fig.l shows a peer-to-peer communication network 100 according to an embodiment of the invention.

The peer-to-peer communication network 100 has a plurality of peers 101, which are arranged according to a ring topology in this embodiment.

The peers 101 may be configured as a personal digital assisstant, a desktop computer or a laptop, etc. The peers each have at least one processor, ^¬ example, a microprocessor, and at least one storage on more.

By means of the peers 101, more precisely in the memories of the peers 101, data are stored. A connecting line 102 Zvi ^¬ rule two peers 101 indicates that the two peers are directly adjacent 101 in the peer-to-peer communication network 100 to each other. The more connecting lines 102 between two peers 101, the less the two peers 101 in the peer-to-peer communication network 100 are adjacent. Conversely, the fewer connecting lines 102 between the two peers 101, the more adjacent two peers 101 in the peer-to-peer communication network 100.

Each peer 101 is a node ID, i. a unique identification within the peer-to-peer communication network 100 assigned. Two peers 101 are then in the peer-to-peer

Peer communication network 100 more adjacent, if their associated node IDs differ little. For example, two peers 101 are then disclosed to each other directly Benach ^¬ when their node IDs differ only in the least sig- nifikanten bit or more least significant bits. As mentioned, the peers 101 serve to store data, for example the peer-to-peer communication network 100 implements a file-sharing communication service for the peers 101. The peer-to-peer communication network 100 is implemented by means of an IP

Communication network, for example by using the In ^¬ ternets realized.

A peer 101 is, as described above, lisiert by an appliance does, for example, a notebook, a PDA (Perso nal ^¬ Digital Assisstant) or a desktop computer. A device can also implement a plurality of peers 101, for example, a plurality of software client units can be executed on a computer system, each realizing a peer 101.

It is assumed that multiple peers 101 have the same user, i. a user operates multiple peers 101. The determination of node IDs for the user-operated peers 101 (and hence their arrangement within the peer-to-peer communication network 100) will now be described.

It is assumed that the peer-to-peer communication network 100 according to the CHORD algorithm be ^¬ is exaggerated. However, the procedure described below is also applicable to a peer-to-peer communication network 100, which is operated in accordance with the Kademlia algorithm and configured according to a tree topology. In general, the peer-to-peer communication network 100 may be operated according to any peer-to-peer algorithm.

2 shows a flow chart 200 according to an embodiment of the invention.

The illustrated procedure is to determine node IDs for peers 101, all operated by the same user. In step 201, a unique name of the user is determined, for example a customer number of the user, his e-mail address, the SIP URI (Session Initiation Protocol - Uniform Resource Indicator) assigned to him or his telephone number.

For example, the customer number of the user 0123456789 is determined as the unique name of the user.

In step 202, a hash value is determined from the unique name of the user using a hash function, for example according to SHA-I or MD5.

For example, from the customer number of the user

(0123456789) the SHAl hash value 87 AC EC 17 CD 9D CD 20 A7 16 CC 2C F6 74 17 B7 IC 8A 70 16 detected.

In step 203, the UUID (unique user identifier) is determined from the determined hash value. In this example, this is done by simply truncating the hash value. It is assumed that the peers 101 in the peer-to-peer communication network are assigned 10-byte node IDs. In this case, the hash value is truncated to 9 bytes, so:

Hash value: 87 AC EC 17 CD 9D CD 20 A7 16 ... User UUID: 87 AC EC 17 CD 9D CD 20 A7

This UUID of the user obtained by chance, the Benut ^¬ is zer but permanently assigned.

In step 204, the user can select node IDs itself ^¬, by complementing the UUID by self-selected bits. In this example, the user augments the UUID by one byte to the 10 bytes used to address the peers 101 in the form of node IDs. For example, the user supplements the UUID in the following manner to determine node IDs for three peers 101 that he or she is operating:

User UUID: 87 AC EC 17 CD 9D CD 20 A7

Node ID for Peer 1: 87 AC EC 17 CD 9D CD 20 A7 01

Node ID for Peer 2:87 AC EC 17 CD 9D CD 20 A7 81

Node ID for Peer 3:87 AC EC 17 CD 9D CD 20 A7 82

The UUID is thus firmly prescribed to the user, but he can freely select the last bits (the last 8 bits in this example) to determine himself node IDs for his peers 101. In this example, the user may use 256 peers 101 using his UUID.

The user-selectable bits can be chosen randomly by the user, for example using a random number generator.

The peers 101 user's different in this from ^¬ management for only the significant least-significant bits (assuming that the signifikantesten bits left ste ^¬ bonds). Thus, the peers 101 of the user are highly contiguous in the peer-to-peer communication network 100. High stability of a peer 101 (ie, a long retention time in the peer-to-peer communication network 100) thus benefits the other peers 101 of the user because it does not generate much load on them (assuming that a high workload always affects strongly adjacent peers 101, which may be the case, for example, if they store the same data).

Conversely, the other peers 101 are overloaded user be ^¬ when he often removed and re-introduced a peer 101 from the peer-to-peer communication network 100th The user is thereby caused to keep his peers 101 as stable as possible in the peer-to-peer communication network 100. In the case of a random distribution of the node IDs of the peers 101 of the user, however, the one of a peer 101 would Entering and leaving the peer-to-peer

Communication network 100 generated random load on any ^¬ peers 101 in the peer-to-peer communication network 100 distribute.

Another advantage is that the peers 101 adjacent to a peer 101 are also likely to be more closely adjacent to the IP network level topology as they are operated by the same user. This has the advantage that keep-alive messages as well as information about changes to the neighbor list (stabilization messages) in the peer-to-peer communication network 100 are exchanged faster and cause less traffic in the entire IP communication network.

Another advantage of this method is that multiple devices of a single-search user based on the user's UUID can be found in the peer-to-peer communication network 100.

However, it may also be a disadvantage if all of the user's devices are adjacent in the peer-to-peer communications network 100. Since the peers 101 are adjacent with high probability even in the IP communications network, they are very likely to all units by means of the same A ^¬ of the IP communication network, for example by means of the same router to the IP communication network connected.

In the event of failure of this router, all peers 101 would be disconnected from the IP communication network simultaneously. In this case, the probability of data loss in the peer-to-peer communications network 100 increases sharply, especially if the number of peers 101 of the user exceeds the number of peers 101 in a replication group (ie, the peers 101, the same data store) reaches or exceeds. However, this problem can be avoided by adjusting the size of the replication groups. Operates a user For example, a maximum of n peers 101 at a time, the replication groups should be increased by (n-1) peers. For a maximum of two peers 101 per user, this approach is quite feasible.

Supplementing the UUID to determine the node IDs at the least significant end of the UUID (i.e., by adding least significant bits) is not desirable, for example because of the associated locality (strong neighborhood) of the user ^¬ surrounded peers 101, so other bits at another location of the UUID may be freely selected.

The closer these arbitrary bits are to the most significant end of the UUID, the more the user's peers 101 are distributed in the peer-to-peer communications network 100.

Thus both the advantages and the disadvantages of the locality of the user's peers 101 are lost.

Nonetheless, multiple peers 101 of the same user are still allowed to exist in the peer-to-peer communication network 100.

As an example, the node IDs are selected by allowing the user to freely select the highest byte of the UUID. A node ID thus results from the freely selected byte and from the following places of the UUID, which in this case is determined by truncating the hash value to a total of 10 bytes.

Example:

Customer number of the user: 0123456789

Hash value of the customer number: 87 AC EC 17 CD 9D CD 20 A7 16 CC 2C F6 74 17 B7 IC 8A 70 16 (20 bytes total)

UUID of the user: 87 AC EC 17 CD CD 9D (cut hash value to 10 bytes abge ^¬) 20 A7 16

Peer 1:

01 AC EC 17 CD 9D CD 20 A7 16 (10 bytes = 80bits)

Peer 2:

80 AC EC 17 CD 9D CD 20 A7 16

Peer 3:

81 AC EC 17 CD 9D CD 20 A7 16

In another embodiment, unique identifiers of the peers 101 determines the user characterized in that the unique identification of the user selectable endings be chosen freely through zer of the Benut ^¬. Clearly, a unique ending is selected before the hash value calculation and included in the hash value calculation.

The node IDs are the unique name of the

Peers 101 is determined by determining the hash values belonging to the unique names of the peers 101 and truncating them to the required length (10 bytes in this example).

Example:

Customer number of the user: 0123456789

Hash value:

87 AC EC 17 CD 9D CD 20 A7 16 CC 2C F6 74 17 B7 IC 8A 70 16

(20 bytes) User's UUID: 87 AC EC 17 CD 9D CD 20 A7 16 (10 bytes)

Unique name of peer 1: 0123456789a

Hash Value: 4C 8E 08 F9 40 44 D9 B5 99 13 F6 68 73 AF 8E AC 42 99 26 6E (20 Bytes)

Node ID of Peer 1: 4C 8E 08 F9 40 44 D9 B5 99 13 (10 bytes)

Unique name of peer 2: 0123456789b

hash:

58 C8 27 05 3B D8 16 F6 BB 5F 00 3E 90 5C 57 D5 88 73 AB 5A (20 bytes)

Node ID of Peer 2:58 C8 27 05 3B D8 16 F6 BB 5F (10 bytes)

By doing so, the peers 101 are randomly distributed in the peer-to-peer communication network 100. The advantages and disadvantages of the location of the peers of 101 Benut ^¬ zers are thus avoided. The following publications are cited in this document [1] Ion Stoica, Robert Morris, David Karger, M. Frans Kaashoek, Hari Balakrishnan: "CHORD: A Scalable Peer-to-Peer Lookup Service for Internet Applications"

[2] Petar Maymounkov, David Mazieres: "A Peer-to-Peer Information System Based on the XOR Metrie"

Claims

claims

1. Peer-to-peer communication device with a memory in which a peer-to-peer identification information of the peer-to-peer communication device is stored, which has a unique unchangeable part and a variable part.

2. peer-to-peer communication device according to claim 1, wherein the unique unchangeable part is a unique Be ^¬ user identification of the user of the peer-to-peer communication device.

3. peer-to-peer communication device according to claim 1 wherein the unique user identification means of a

Hash function is determined from a unique name of the user.

4. Peer-to-peer communication device according to claim 3, wherein the unique user identification is a telephone number, an e-mail address, a SIP URI or a customer number of the user.

5. Peer-to-peer communication device according to one of the claims 1 to 4, wherein the variable part is one of the Be ^¬ user of the peer-to-peer communication device freely selectable bit combination.

6. Peer-to-peer communication device according to one of the claims 1 to 5, wherein the peer-to-peer

Communication device is part of a peer-to-peer communication network, which is operated according to the CHORD algorithm or according to the Kademlia algorithm.

7. peer-to-peer communication device according to one of claims 1 to 6, wherein the peer-to-peer communication device by means of a personal digital assistant, a desktop computer or a laptops reali ^¬ Siert.

8. Method for forming a peer-to-peer identification for a peer-to-peer communication device, in which the peer-to-peer

Identification is formed of a unique unchangeable part and a variable part.

9. Method for Operating a Peer-to-Peer Communication Network Having Multiple Peer-to-Peer

Communication devices, in which the peer-to-peer communication devices are addressed in each case by means of a peer-to-peer identification information, which has a unique unalterable ^¬ part and a variable part.