Chinese Journal of Electronics Vol.24, No.4, Oct. 2015

Research on the Model of a Lightweight Resource Addressing∗ LUO Bingqing and SUN Zhixin (College of Internet of Things, Nanjing University of Posts and Telecommunications, Nanjing 210003, China) Abstract — This paper discussed the characteristics of addressing from the perspective of Internet addressing mechanism. An Internet of things (IOT) resource addressing iteration model was defined. In the model, a direct addressing mode for active nodes and an indirect addressing mode for passive codes were proposed, which meet the requirement for multiple encoding mode. A unified IOT resource lightweight addressing scheme based on IPv6 has been proposed to implement the two addressing modes. The scheme utilized the virtual domain to solve the problem of the heterogeneous encoding. The paper implemented the addressing process from the Internet host to the sensor node based on IPv6 over low-power wireless personal area networks (6LoWPAN) protocol. The experiment results show that the scheme is performed to realize communication between wireless sensor networks and IPv6 networks. Key words — Addressing, Lightweight, IPv6 networks, Iteration model, Unified addressing scheme.

I. Introduction With the maturity of the Internet addressing scheme and the development of IOT, the addressing and coding mechanism of IOT has become an inevitable problem in the future. As an extension of the Internet, the addressing mode of IOT needs to be associated with the pattern of Internet. At present the addressing mode of the Internet includes: addressing between domain name and IP address[1] ; addressing between IP address and MAC address[2] ; addressing between IP address and IP address[3] . The research of addressing and coding mechanism for IOT aims at realizing the interconnection between terminals in IOT and hosts in the Internet. The most serious problem of addressing and coding mechanism for the IOT is the unimplemented general technical standard. The academic research of IOT is divided into local area research[4,5] and wide area research. However, the terminal nodes in the local area network

could not communicate with the hosts in the Internet directly, which could not meet the requirement of addressing for IOT. As a consequence, the problem of addressing mode of wide area network has aroused some researchers interest. Ref.[6] proposed an IPv6 addressing mode using simplified IPv6 coding method. Ref.[7] proposed a low-power address auto configuration and routing redirection mechanism based on Ref.[6] to reduce the cost of terminal. From the perspective of protocol, Toumi K. put forward a HIP Address protocol called HAT[8] to realize the conversion of Radio frequency identification (RFID) code to IP address. While not all the devices are equipped with the ability to run on the IPv6 protocol stack, the above study also failed to give full implementation of addressing scheme from the sensing node to the Internet host. Refs.[9,10] focused on all-IP communication between wireless network and IPv6 network, but the papers did not provided a communication scheme for the devices which could not equipped with IPv6 address. Ref.[11] put forward an address translation mechanism to realize the communication between nodes residing in and residing outside the IEEE 802.15.4 network. In the proposal, the IDs allocated to the external nodes should be unique and unified which is unrealistic. Ref.[12] proposed a global-tolink-layer address translation at the gateway, while the paper also failed to consider the devices without link layer address. Ref.[13] expanded the layered iteration model of the Internet resource naming and addressing, and proposed a general layered model of IOT addressing. However, the model is only limited to things with coding terminal, the description of addressing objective is ignored. For the most important, the implementation of the model needs a lot of external condition to support, so the feasibility of the scheme is to be discussed. The paper proposes an IOT resource addressing iteration model under multi encoding, and the paper has the following contributions: 1) An IOT resource addressing iteration model is proposed based on the characteristics of the Inter-

∗ Manuscript Received Dec. 12, 2013; Accepted Jan. 12, 2014. This work is supported by the National Natural Science Foundation of China (No.60973140, No.61170276, No.61373135), the Research Project of Jiangsu Province (No.BY2013011), the Jiangsu Provincial Science and Technology Enterprises Innovation Fund Project (No.BC2013027), the High-level Personnel Project Funding of Jiangsu Province Six Talents Peak and Jiangsu Province Blue Engineering Project, and the Major Project of Jiangsu Province University Natural Science (No.12KJA520003) c 2015 Chinese Institute of Electronics. DOI:10.1049/cje.2015.10.028 

Research on the Model of a Lightweight Resource Addressing net resource addressing scheme, In the model, active nodes and passive nodes have been classified and defined comprehensively, and direct addressing mode and indirect addressing mode are proposed based on the type. 2) Based on the model, a unified IOT resource lightweight addressing scheme based on IPv6 has been proposed to implement the two addressing modes. The scheme utilizes the virtual domain to solve the problem of the heterogeneous encoding and raises RFID repeaters and 6LoWPAN coordinator to realize the indirect addressing mode for nodes that have weak processing ability or do not have processing ability. 3) The paper implements the addressing process from the Internet host to the sensor node based on 6LoWPAN protocol proposed by Internet engineering task force (IETF)[14,15] .

II. IOT Resource Addressing Iteration Model Differs from Internet resources, the collection of IOT resources is not only from the sensor nodes, but also from many objects with object coding. Therefore, the research on addressing scheme of IOT is divided into two parts, namely the addressing mode and the code resolution. So the realization of unified resource addressing iteration model of IOT should be based on the unification of coding diversity. In the sensing layer of IOT structure, the resource includes two kinds of terminal: active nodes and passive nodes. Active nodes: Refers to the nodes with processing ability, IP address and MAC address, such as sensor nodes, cameras, mobile terminals and so on. Passive nodes: Refers to the nodes without processing ability, and with a tag to store the nodes information, such as objects with RFID tags. In the sensing layer of IOT structure, the maturity coding mode includes RFID terminal code, MAC code, compressed IP code and other short address code. They are all direct encoding except the compressed IP code. Direct encoding D: Refers to be a particular entity, which lies at the bottom of interactive addressing process, directly identify some entity, and is locally unique, such as MAC address in the Internet. Indirect encoding MN : Refers to be the medium in the iterative addressing process, indirectly identify some entity. An entity may have multiple indirect encoding, such as IP address in the Internet. Similar to the Internet addressing mechanism, the addressing scheme of the IOT shows the hierarchical iteration, among which the iteration means that get an indirect encoding in an addressing process based on the objective encoding, and then make this indirect encoding to be the input of the second addressing process. Every addressing process is a layer of the whole resource addressing scheme, and belongs to an independent subsystem. The resource addressing system only needs to know the interface information of different coding scheme without needing to understand the logic in all layers, which reduces the coupling between each addressing system. Therefore, the implementation of the addressing scheme could focus on the implementation of every subsystem instead of concentrating on

833

the interaction between different layers. The implementation of resource addressing system in each layer belongs to different addressing layers in logic no matter uses the same addressing system or not. The output of an addressing subsystem could be used by many lower layers, and an addressing subsystem could be called by its upper layer resource addressing subsystem[13] . Fig.1 shows IOT resource addressing iteration model. Direct addressing mode Refers to the addressing mode which the termination of the iteration is the direct encoding of IOT resource. Indirect addressing mode Refers to the addressing mode which the termination of the iteration is the direct encoding of superior resource. Superior resource In logic, one resource A is administered by another resource B. The administer B is the superior resource of the administered resource A; or the information of one resource A always transfers to its logical upper resource B by the resource A, then the upper resource B is the superior resource of the resource A. In Fig.1, indirect encoding N is the input of the addressing subsystem N, and the indirect encoding N +1 is the output. In the meanwhile, the indirect encoding N +1 is the input of the layer N +1. The layer N and the layer N +1 are relative, but the layer N has no coupling relationship with the layer N +2. If the end of the iteration is the direct encoding of active nodes, then the addressing process is over, and the addressing mode is direct addressing. If an active node is the superior node of a passive node, then the direct encoding analysis information could be obtained through Resource information positioning mechanism (RIPM) in layer N +3.

Fig.1. IOT resource addressing iteration model

Assume the indirect encoding is MN and the direct encoding is D, the indirect encoding space in layer N is shown in Eq.(1). The direct encoding space in layer N is shown in Eq.(2). N ameSpaceMN = {M1 , M2 , ..., Mj , ..., Mk }

(1)

N ameSpaceD = {D1 , D2 , ..., Mj , ..., Dk }

(2)

Define the addressing function of resource addressing system N is the unary function ASN , the Eq.(1) could be charged into Eqs.(3) and (4). N ameSpaceM(N +1) = ASN (N ameSpaceMN )

(3)

Mi = Mj ⇒ AS(Mi ) = AS(Mj )

(4)

Chinese Journal of Electronics

834

The process of direct addressing scheme is shown in Eq.(5). N ameSpaceD = ASN (ASN−1 (ASN−2 ...

2015 

the Internet through a specific LBR, expressed as L {l1 , l2 , l3 , ...ln , g} , where g is the LBR.

=

(ASK+1 (ASK (N ameSpaceMK ))))) (5) And 1 ≤ K ≤ N ≤ M . The IOT resource addressing iteration model proposed by this paper has the following characters. 1) The proposed addressing modes can be applied in both of the sensor nodes and the object tags without processing ability. 2) The direct addressing mode and the indirect addressing mode are proposed to meet the different requirements of terminal resources. The model is compatible and expandable. The direct addressing mode is suitable for the addressing scheme of the Internet, the terminals in IOT could be interconnected with the hosts in the Internet easily; while the indirect addressing mode could satisfy the requirements of multi encoding through RIPM.

III. A Unified IOT Resource Lightweight Addressing Scheme Based on IPv6 According to the IOT resource addressing iteration model proposed in Section II, a unified IOT resource lightweight addressing scheme based on IPv6 will be shown in this section. As the main encoding of the sensor nodes, IPv6 address has the following advantages. 1) The terminals with IPv6 address could interconnect with the hosts in the Internet easily. 2) IPv6 address auto-configuration is an important technical feature of IPv6 and it can allocate a unique IPv6 address for each terminal, which is consistent with the design goals of Wireless sensor network(WSN), such as self-organization and self-configuration, etc.[16] Fig.2 shows the IOT resource lightweight addressing structure based on IPv6, which is consist of sensor nodes, the LoWPAN border router (LBR), resource manager system and some application servers. In the sensing layer of IOT structure, the terminal nodes are divided into different virtual domains according to their different encodings or protocols. Differs from the concept of gateway in Internet, LBR is not only equipped with the ability of package forwarding, but also the ability of address conversion between different sensor networks. Considering the diversity, heterogeneity, and large quantity of IOT terminals, the resource manager system is constructed to locate and access the terminal resource effectively with resource addressing scheme. 1. Related definition 6LoWPAN node Refers to the nodes with 6LoWPAN protocol stack, every 6LoWPAN node could send and receive IPv6 packets with an unique IPv6 address, as well as support ICMPv6 and UDP protocol. 6LoWPAN domain Refers to WSN which includes N 6LoWPAN nodes, expressed as L={l1 , l2 , l3 · · · ln }. Every 6LoWPAN domain is consisted of the 6LoWPAN nodes with the same IPv6 prefix (64 bit). The IPv6 address remains the same no matter what is the position of nodes in the domain. Simple LoWPAN domain Refers to a simple low cost communication network, where the sensor nodes connect with

Fig. 2. The IOT resource lightweight addressing structure based on IPv6

Expanded LoWPAN domain Refers to a simple low cost communication network,where the sensor nodes connect with the Internet via multiple LBRs, expressed as  L ={l1 , l2 , l3 · · · ln , g1 , g2 , · · · gn }, gi is the LBR. Tag domain Refers to a wireless sensor network, where the sensor nodes labeled by N tags, expressed as S= {s1 ,s2 ,s3 · · · sn } . The wireless sensor network based on multiple protocols expressed as I, and the wireless sensor network is neither the 6LoWPAN domain nor the tag domain expressed as O = I ∩ L ∪ S. In gather O , terminal wireless transmission protocol of IOT could be ZigBee, Bluetooth, WIFI and so on. RFID repeater Refers to a RFID interrogator with an unique IPv6 address. 2. Direct addressing mode In Fig.3, all the sensor nodes are active nodes in 6LoWPAN domain (including simple LoWPAN domain and expanded LoWPAN domain). These nodes have their unique IPv6 address and could be generally addressed by direct addressing mode. Taking ping test as an example, a typical lightweight direct addressing mode based on IPv6 is shown in Fig.3. In Fig.3, node A is a host in the IPv6 network, node B is a sensor node with a domain name iot.test.com. Assuming that context is known, do the operation A PING6 B: ping6 iot.test.com according to the following process. 1) The parameter of the command PING6 is a domain name, the host A send a Domain Name System (DNS) request to DNS server, the domain name of the node B is iot.test.com. 2) Do addressing through DNS, getting the IPv6 address of iot.test.com: 1080::4:123:275F:893D. 3) DNS server sends the IPv6 address 1080::4:123:275F:893D to the host A. 4) The host A generates a ICMP ACK request packet, and the head of the IPv6 packet is standard IPv6 head, the destination address is the host A s address as well as the source address is the node B  s address. 5) After the host A receives the ICMP ACK request packet, it sends the ICMP ACK request packet to LBR. 6) After LBR receives the ICMP ACK request packet through ad-

Research on the Model of a Lightweight Resource Addressing dressing scheme in network layer, the head of IPv6 packet will be compressed by 6LoWPAN adapter protocol, which could make the packet conform to the IEEE 802.15.4 standard. 7) After the compressed, LBR sends the compressed ICMP ACK request packet to the node B. 8) The node B receives the compressed ICMP ACK request packet through addressing scheme in link layer, then the node B extracts IP packet from the frame. Meanwhile, the standard IPv6 head will be uncompressed to a standard IPv6 head by 6LoWPAN adapter protocol. 9) The node B generates a ICMP ACK reply packet after the analysis. The head of the IPv6 packet is a standard IPv6 head, and the destination address is the host A s address as well as the source address is the node B  s address. 10) The standard IPv6 head in 9) is compressed by 6LoWPAN adapter protocol. 11) The host B sends the compressed ICMP ACK reply packet to LBR. 12) After LBR receives the compressed ICMP ACK reply packet, which will be uncompressed by 6LoWPAN adapter protocol to conform to the Internet standard. 13) LBR sends the ICMP ACK reply packet to the host A through addressing scheme in network layer. 14) Finally, when the host A receives the ICMP ACK reply packet, the whole addressing process is over.

835

responsible for allocating a pair of ZigBee address and IPv6 address to nodes in its administration ZigBee network and the destination nodes communicated in the network, in this way 6LoWPAN coordinator is transparent to the ZigBee network. The nodes in tag domains are passive nodes proposed in Section II. Through the way of addressing the direct encoding of superior resource, the tag resource information would be found according to the resource information position mechanism. In the lightweight addressing structure of IOT based on IPv6, the superior resource of the nodes in tag domains is RFID repeater, which could work with uIPv6 protocol stack. The RFID repeater as the superior repeater of the nodes in tag domains allocates IPv6 address to the nodes and executes the indirect addressing mode for RFID terminal.

IV. Implementation To test the feasibility of the lightweigh addressing structure based on IPv6, a real network deployment was to be carried out as shown in Fig.4. In the experiment, a sensor attached to the computer via USB provided the ability of sending and receiving 6LoWPAN packets as the LBR. Another sensor node was acted as a sensing node to communicate with sink nodes. A remote client server was also needed in this experiment; at last there was another computer as the IPv6 host communicated with the LBR via the Internet.

Fig. 4. Experiment structure

Fig. 3. Ping process of lightweight direct addressing mode based on IPv6

The IPv6 protocol stack is uIPv6 protocol stack, which is different from the standard IPv6 protocol stack. In addition, in the above process, the node B and LBR is assumed to connect directly. But in the actual environment, the packets sent by LBR may be through multiple hops to reach the destination node, and packet routing and forwarding are completed by relay node. 3. Indirect addressing mode The wireless transmit protocol may be ZigBee, Bluetooth, IEEE 802.15.1 or WIFI for those other sensor network domain in the lightweight addressing structure of IOT. As shown in Fig.2, for the nodes in these domains, the paper raised a proposal that uses 6LoWPAN coordinator to connect these nodes to LBR. In the case of ZigBee, the 6LoWPAN coordinator is

1. LBR ping the sensor node Sensor node with IPv6 address belongs to active node, which can interact with LBR through the way of direct addressing. In order to verify the implementation model introduced in Section III, a IPv6 address of remote node was measured in advance, which is expressed as 2001:acf8:42ed:2590: 212:7400:1467:ac69. This address is indirect encoding in addressing model proposed by this paper, that is the M in Eq.(1). By iterating the Eq.(5), directing encoding of the node, which is also the MAC address of the sensor node, could be obtained. Fig.5 shows the result of PING6 2001:acf8:42ed:2590: 212:7400:1467:ac69.

Fig. 5. The LBR ping the sensor node

Chinese Journal of Electronics

836

The experiment result verifies the interconnection between the LBR and the sensor node based on IPv6. Operation PING is a typical process of direct addressing, while temperature and humidity information sensed by wireless sensor nodes can also be collected by LBR through the process of direct addressing. 2. The Internet hosts collect sensor data On the basis of successful PING test, a temperature and humidity data collection test was deployed in this section. The encoding analysis process is the same to the Internet addressing; although the encoding of the sensor node is IPv6 address, the addressing process is still based on the transformation between IP address and MAC address. The experimental result is shown in Fig.6. The connection between the host and sensor nodes was established through the way of network initialization. After that the sensor node sends the temperature and humidity data in every 4 seconds.

[4]

[5]

[6]

[7]

[8]

[9]

[10]

[11]

[12] Fig. 6. The host in the Internet reads the sensor data

A direct addressing process between the Internet host and the sensor node in WSN has been completed by programming in the application layer. The experiment result has verified the lightweight addressing scheme based on IPv6 is in favour of the interconnection between WSN and the Internet.

[13]

[14]

[15]

V. Conclusion The paper proposes a lightweight unified addressing model for multi encoding to achieve the interconnection between WSN and the Internet based on IPv6. The paper analyzes and experiments the model’s implementation feasibility and the experiment data shows the interconnection between WSN and the Internet based on IPv6. At the present time, a unified addressing scheme between WSN and IPv6 networks is a very popular research field but the relative study is still in the initial stage. Therefore, we hope that the paper can contribute to the future research studies. References [1] X. Chen, H. Wang, S. Ren, et al., “Maintaining strong cache consistency for the domain name system”, Knowledge and Data Engineering, Vol.19, No.8, pp.1057–1071, 2007. [2] S.Y. Nam, D. Kim, J. Kim, et al., “Enhanced ARP: Preventing ARP poisoning-based man-in-the-middle attacks”, Communications Letters, Vol.14, No.2, pp.187–189, 2010. [3] G. Raja and M. Elad., “RIP-based near-oracle performance

[16]

2015

guarantees for SP, CoSaMP, and IHT”, IEEE Trans.Signal Processing, Vol.60, No.3, pp.1465–1468, 2012. L.J. Song, J.P. Wu, Y. Cui, et al., “A novel load balancer for scalable Multi-NAT”, Chinese Journal of Electronics, Vol.22, No.4, pp.712–716, 2013. Y.G. Shu, Z.Y. Shu and B. Luo, “An energy balancing adaptive routing strategy based on interference model in 6LoWPANs”, Chinese Journal of Electronics, Vol.22, No.3, pp.466–470, 2013. J.I. Agbinya and M.A. Agbinya, “A new address allocation scheme and planning of personal area and sensor networks”, International Symposium on Parallel and Distributed Processing with Applications, Sydney, Australia, pp.463–470, 2008. Z. Jun,Y. Bo and M. Aidong, “Address allocation scheme of wireless sensor networks based on IPv6”, 2nd IEEE International Conference on Broadband Network & Multimedia Technology, Beijing, China, pp.597–601, 2009. K. Toumi and M. Ayari, “HAT: HIP address translation protocol for hybrid RFID/IP Internet of things communication”, International Conference on Communication in Wireless Environments and Ubiquitous Systems: New Challenges, Sousse, Tunisia, pp.1–7, 2010. X.N. Wang and S. Zhong, “All-IP communication between wireless sensor networks and IPv6 networks based on location information”, Computer Standards & Interfaces, Vol.35, No.1, pp.65–77, 2013. X.N. Wang and H.Y. Qian, “Research on all-IP communication between wireless sensor networks and IPv6 networks”, Computer Standards & Interfaces, Vol.35, No.4, pp.403–414, 2013. D.I. Choi, J. Park, S.Y. Kim,et al., “Improve IPv6 global connectivity for 6LoWPAN”, Advanced Communication Technology, Seoul, Korea, pp.1007–1010, 2011. A. Zimmermann, et al., “6glad: IPv6 global to link-layer address translation for 6lowpan overhead reducing”, Next Generation Internet Networks, Krakow, Poland, pp.209–214, 2008. N, Kong, X.D. Li, W.M. Luo, et al., “Model of the resource addressing in the Internet of Things”, Journal of Software, Vol.21, No.7, pp.1657–1666, 2010. N. Kushalnagar, et al., “IPv6 over Low-power wireless personal area networks (6LoWPANs): Overview, assumptions, problem statement, and goals”, RFC, 4919, 2007. G. Montenegro, N. Kushalnagar, J. Hui, et al., “Transmission of IPv6 packets over IEEE 802.15.4 networks”, RFC, 4944, 2007. X.N. Wang and H.Y. Qian, “An IPv6 address configuration scheme for wireless sensor networks”, Computer Standards & Interfaces, Vol.34, No.3, pp.334–341, 2012. LUO Bingqing was born in Ju Rong, Jiangsu Province, in 1987. She is now a Ph.D. candidate in the college of Internet of Things, Nanjing University of Posts and Telecommunications.Currently her research interests include IP wireless sensor network, wireless sensor network communication, mobile Internet, etc. (Email: [email protected])

SUN Zhixin was born in Xuan Cheng, Anhui Province, in 1964. He is now a professor and Ph.D. supervisor of the college of Internet of Things in Nanjing University of Posts and Telecommunications.Currently his research interests include computer networks and security, multimedia communication, mobile Internet, etc. (Email: [email protected])