Cellular-connected unmanned aerial vehicles (UAVs) are increasingly being deployed in emerging Internet of Things (IoT) applications, where reliable handover management is critical to ensure uninterrupted communication. Unlike terrestrial users, UAVs face frequent handovers due to antenna side-lobe coverage and fragmented aerial cell overlaps. Existing optimization strategies, however, are often limited to coarse-grained “whether-to-handover” decisions or unidirectional cell individual offset (CIO) tuning, resulting in degraded handover performance. To address these problems, this paper proposes an Event-triggered Bidirectional CIO Optimization with Hybrid prioritized experience replay based Deep Reinforcement Learning (EBCO-HDRL) algorithm. Under the A3 event-triggered mechanism, EBCO-HDRL jointly configures serving-to-neighbor and neighbor-to-serving CIOs, improving handover accuracy and reducing ping-pong events. An event-triggered optimization strategy adaptively reconfigures CIOs only when needed, mitigating computational overhead and enhancing training stability. To further improve sample efficiency, a Hybrid Prioritized Experience Replay (HPER) scheme is introduced, combining temporal-difference error and high-reward sampling, supported by a heterogeneous neural network design. Simulation results show that EBCO-HDRL significantly outperforms baseline algorithms in terms of precision, stability, and robustness, offering an effective solution for UAV handover management in cellular IoT networks.

Semantic Edge Segmentation (SED) is crucial for intelligent agents to understand and interact with their environments, as it enables them to locate and recognize semantic boundaries. The prevailing framework in the field of SED is multi-label learning, which identifies edges and their semantics by learning to assign multiple labels that indicate the categories of the objects forming the edges. However, this framework has demonstrated limited performance when dealing with complex scenarios. In this paper, we propose a mask classification framework specifically tailored for the SED task, termed EdgeMaskFormer. Within this framework, we develop a query-based edge semantic extractor to learn semantic embeddings for edge mask classification with assistance from regional semantic supervision. Additionally, we design a context-aware hierarchical edge extractor to serve as an edge mask head, which can capture multi-scale edges of different categories under the guidance from the semantic embeddings via dynamic convolution. Furthermore, we develop matching and supervision mechanisms specifically for edge mask classification in order to reduce edge noise and address the imbalance between edge and non-edge samples. Our extensive experiments on three public datasets demonstrate that the proposed approach achieves outstanding performance in semantic edge detection, particularly on those datasets with complex scenarios.

The existing Internet treats all packets equally, which means that a packet can be dropped entirely due to network congestion. To address this issue, a new packetization scheme and transport paradigm called Qualitative Communication, along with its associated packet wash operation, have been proposed. Qualitative Communication is particularly beneficial for transmitting multimedia content in the presence of network congestion. This paper specifically focuses on the transmission of JPEG/MPEG packets and investigates the characteristics of the Discrete Cosine Transform (DCT) used in the JPEG/MPEG encoding process. The proposed encoding scheme organizes the quantization coefficients of all image blocks in a packet based on their significance for image decoding and reconstruction. Additionally, the paper explores the idea that end users may be willing to accept a certain level of quality degradation rather than receiving nothing at all within an expected time frame. The benefits provided by the proposed mechanisms are discussed, and simulation results validate the improvements in the successful packet delivery ratio and the end user's level of satisfaction.

Satellite networking has been a very active research area in both academia and industry recently. Starlink provides high-speed, low-latency Internet access. More companies have claimed to provide similar services soon. It is expected that different industries will embrace satellite network business in the future. This paper reviews the achievements in satellite-related works in standard organizations (IETF and 3GPP) from the networking perspective (Layer 2/3 and above). The issues and challenges in different areas are analyzed, such as IP networking for massive satellite constellations, wireless network architecture for Non-Terrestrial Network (NTN) integration, etc. Based on the latest developments, the paper envisions the evolution direction and outlines the study items for satellite networks in 6G.

Knowledge graph embedding aims to transform the entities and relations of triplets into the low-dimensional vectors. Previous methods are oriented towards the static knowledge graphs, in which all entities and relations are assumed to be known and only some unknown triplets need to be predicted. However, the real-world knowledge graphs can grow dynamically, and some new knowledge are often added. To embed the new knowledge into the space of original knowledge graph, the classic models have to perform the entire re-embedding with including the new and original knowledge. This causes heavy computational burden for embedding. To address this problem, this study proposes a new model of anchors-based incremental embedding (ABIE) to implement the dynamical embedding for the growing knowledge graph. According to ABIE, every knowledge graph has some key entities, called anchors, which can fix the embedding space of knowledge graph. When some new knowledge is added into the graph, only a few updated entities and relations are embedded into the embedding space with the help of anchors, and the entire re-embedding on the whole graph is not necessary. By this way, the computational burden of embedding caused by the growth of knowledge graph is reduced significantly.

Access control is one of the most basic techniques to ensure the security of the information system. The traditional access controls of information systems are usually performed based on the traversals or queries of rules. However, with the increasing complexity of information systems, massive data, and open environments bring great workload and risk for the traditional methods. This study proposes a model of embedding-based computable access control (ECAC), by employing the idea of representation learning in artificial intelligence. According to ECAC, access control rules can be embedded into a Euclidean vector space, and the security of arbitrary behavior can be computed by numerical vector operations, without any traditional querying or traversing of rules, and thus the workload of access control is reduced. Furthermore, by the embedding-based computation, the security of unknown behaviors can be predicted. Potentially, due to the use of numerical vectors instead of traditional semantic symbols, the risk of privacy leakage via semantics can be reduced. Finally, as the first embedding-based access control model, the effectiveness of ECAC is evaluated and concluded by the experiment-based analyses and discussions.

Stringent end-to-end latency guarantee service has been foreseen to be required by many emerging applications. With the traffic driven by mobile devices continuing to dominate the Internet, end-to-end latency guarantee needs to take into account of the latency incurred on the Internet as well as in the mobile networks. In this paper, we consider the latest 5G cellular network and its architecture, design a special type of slices namely, Latency Guarantee Service (LGS) slice. A unified 3GPP data plane protocol stack based on New IP is also introduced.

The Information and Communication Technology (ICT) industry constitutes roughly 5-9% of the overall energy consumption, and may quickly grow to as high as 20%. With the global warming challenge, the ICT sector sets its net-zero emission goal to be achieved by 2050. This paper discusses how newly proposed New IP framework and its three major components: Shipping Specification, Contract Insertion, Qualitative Communication, support green Internet and reduce energy consumption in IP routers. This paper also summarizes Future research opportunities to enhance the reduction of network power consumption based on the New IP framework.

Precise end-to-end latency guarantee is predicted to be required by many emerging applications. On the other hand, the network traffic will continue to be dominated by mobile devices. Therefore, the end-to-end latency is composed of the latency incurred in the Internet as well as in the mobile networks. In this paper, we target to address the end-to-end latency guarantee requirement for downlink traffic by leveraging the previously proposed 5G slice namely, Latency Guarantee Service (LGS) slice. The mechanisms and procedures are proposed by taking the compatibility of 5G architecture into consideration. The simulation results show that the downlink flows which are admitted by the LGS slices are verified to satisfy the end-to-end latency constraint consistently.

Precise end-to-end latency guarantee is predicted to be required by many emerging applications. On the other hand, the network traffic will continue to be dominated by mobile devices. Therefore, the end-to-end latency is composed of the latency incurred in the Internet as well as in the mobile networks. In this paper, we target to address the end-to-end latency guarantee requirement for downlink traffic by leveraging a new type of 5G slice namely, LGS (Latency Guarantee Service) slice. The mechanisms and procedures are proposed by taking the compatibility of 5G architecture into consideration. The simulation results show that the downlink flows which are admitted by the LGS slices could satisfy the end-to-end latency constraint consistently.

Autonomous driving is complemented by remote driving when it falls short. Remote driving requires the end-to-end transmission latency between the vehicles and cloud servers to be extremely small and within a desired deadline, which gives the drivers enough time to respond to emergent situations. This paper focuses on the multi-camera assisted remote driving scenarios, in which images captured by multiple cameras are concatenated and encapsulated in one jumbo packet for overhead reduction and energy efficiency at router side. Jumbo packets are more sensitive to network congestion since they might be completely lost in the current Internet implementation. Qualitative Communication is a novel transport paradigm, that reduces the treatment and processing of packets in the network into even finer granularity. The paper illustrates how Qualitative Communication along with other components of New IP could improve the multiple-camera assisted remote driving experience. Some preliminary results show that the jumbo packet delivery success ratio can be significantly increased.

Different industry verticals require their network domains to describe their optimized application control and behavior through custom protocols. Such local protocols are necessary because adapting specific behaviors using current general-purpose network paradigms often leads to constrained design decisions since the IP does not provide necessary customization with its existing format. At the same time, with the growth in Industrial Internet and IoT, there is a need for convergence between the standard Internet and proprietary protocols. In this regard, network addresses are key to representing application semantics or a device's capability, but the current IP address structure does not sufficiently capture these artifacts. We propose the use of New IP shipping specification to accommodate a diverse set of local protocol requirements in a well-structured user-defined address scheme and its effective integration and adaptation to the IP for opaque connectivity over the Internet.

The paper proposes knowledge inference and reasoning functionalities to be supported in a distributed manner by network nodes. The recently thriving Internet framework, i.e., New IP could facilitate the enabler of the proposed functionalities with in-network intelligence. The paper describes the new knowledge reasoning request and response procedure with an IoT use case. The performance improvement over the legacy approach is analyzed and illustrated with different parameter settings. It is proved that the proposed in-network knowledge reasoning approach could significantly decreases the average number of nodes that the requests travel in the network, resulting in less waiting time for requesters. On the other hand, there is barely any increase on the message overhead.

Not all data units carried over a packet stream are equal in value. Often some portions of data are more significant than others. Qualitative Communication is a paradigm that leverages the ‘quality of data’ attribute to improve both the end-user experience and the overall performance of the network specifically when adverse network conditions occur. Qualitative Communication allows that the content that is received differs from the one that is transmitted but maintains enough of its information to still be useful. This paper describes Qualitative Communication, its packetization methods, and the corresponding mechanisms to process the packets at a finer granularity with New IP. The paper also discusses its benefits to the emerging and future network applications through a couple of use cases in video streaming, multi-camera assisted remote driving, ARNR and holographic streaming, and high precision networking. Some preliminary performance results are illustrated to show such benefits. This suggests that Qualitative Communication will find wide applications for many use cases.

Accurate end-to-end latency guarantee service has been foreseen to be required by many emerging applications. Today's Internet with the default Best Effort service, and some more advanced services such as IntServ, DiffServ, traffic engineering cannot ensure the latency guarantee service to be precisely offered. This paper proposes to leverage New IP, which is a newly proposed and vigorously developing Internet framework, in order to achieve in-band signaling for flow admission and resource reservation at intermediate routers. The corresponding New IP packets for latency guarantee service setup and feedback are presented. In-band signaling enhances DiffServ to achieve the granular flow-level latency guarantee service with scalable control plane and data plane. There are minimal changes to the existing DiffServ capable hardware. The simulation results show many performance benefits of the proposed schemes: the latency and jitter are bounded with predictable values; the link utilization can be maximized without bandwidth waste; different classes of flows and different flows within the same class can be served fairly without discrimination.

Precise Quality of Service guarantee, especially stringent latency (in-time) guarantee for packet delivery is demanded by many emerging applications, such as remote surgery, autonomous driving, etc. However, today's Internet's best effort and statistical multiplexing do not assure this type of network services at packet level. In this paper, we take advantage of the contract component proposed in the New Internet Protocol (New IP) framework, a packet that requires end-to-end in-time guarantee can be informed to each forwarding network node. Given there might be multiple latency-sensitive packets for the same output port that need to be scheduled, an optimal scheduling algorithm (TDMS) is proposed to achieve minimal average dwell time at each router during packet forwarding. The supporting theorems are proved to verify the simplicity and validity of the proposed algorithm. The simulation results also confirm that the proposed TDMS achieves the best performances on packet in-time delivery and average packet dwell time in routers.

Qualitative Communication has been proposed to reduce the granularity of packet dropping due to congestion, from packet level to chunk level, which can significantly improve the user experienced latency and reduce the number of packet retransmissions under the reliable transmissions. This paper proposes an approach to achieve Qualitative Communication by applying random linear network coding to chunks in packet payload. With the help of New IP framework, the network can efficiently perform packet wash in order to retain as many chunks in the packets as possible when congestion happens. The paper proposes several advanced features to further boost the benefits of this approach.

The IP is a primary data plane protocol on the Internet, which has several deficiencies when addressing the needs of modern digital society involving machine-to-machine communication and a remarkably enhanced user experience. New IP is an advanced network protocol specification to modernize the network layer without changing the fundamental Internet architecture. New IP envisions a new header format with 3 functional characteristics, i.e., shipping spec, contract spec, and payload spec. Using these fundamental blocks, New IP proposes a new data plane forwarding paradigm with far more advanced capabilities, such as ManyNets addressing, high precision services and qualitative communications.

Relation Extraction is concentrated on finding the unknown relational facts automatically from the unstructured texts. Most current methods, especially the distant supervision relation extraction (DSRE), have been successfully applied to achieve this goal. DSRE combines knowledge graph and text corpus to corporately generate plenty of labeled data without human efforts. However, the existing methods of DSRE ignore the noisy words within sentences and suffer from the noisy labelling problem; the additional knowledge is represented in a common semantic space and ignores the semantic-space difference between relations and entities. To address these problems, this study proposes a novel hierarchical attention model, named the Bi-GRU-based Knowledge Graph Attention Model (BG2KGA) for DSRE using the Bidirectional Gated Recurrent Unit (Bi-GRU) network. BG2KGA contains the word-level and sentence-level attentions with the guidance of additional knowledge graph, to highlight the key words and sentences respectively which can contribute more to the final relation representations. Further-more, the additional knowledge graph are embedded in the multi-semantic vector space to capture the relations in 1-N, N-1 and N-N entity pairs. Experiments are conducted on a widely used dataset for distant supervision. The experimental results have shown that the proposed model outperforms the current methods and can improve the Precision/Recall (PR) curve area by 8% to 16% compared to the state-of-the-art models; the AUC of BG2KGA can reach 0.468 in the best case.

When a content is published in the Internet, it is usually associated with many semantics information, which can be used to identify and discover the content. The current Internet protocols do not support the semantics aware content request very efficiently. By taking advantage of the recently developed Big IP Protocol (BPP), which brings the intelligence into the network while still maintaining the traditional communication paradigm, we propose in this paper a novel semantics and deviation aware content request scheme. It considers the possibility that a consumer may be able to accept the semantics information of the received content having certain deviation from his/her requirements. The paper proposes the structures of BPP enabled content request and content messages, as well as the detailed procedures taken by the network nodes to process those messages. The performance analysis shows that the latency experienced by the consumer is reduced tremendously compared to the traditional approach, with the extra overhead to be reasonable small in different settings and scenarios.

The paper proposes the solution to guarantee the extremely low latency or even zero latency to the moving subscribers in requesting the multimedia streaming service. The solution leverages the competent network resources that could be provided by the 5G network and constructs a slice with appropriate resource allocation on a per user basis. The solution is context aware, including user's location, travel route, driving speed, and road traffic condition. The network resource allocation algorithms are proposed to make sure that the storage and bandwidth are appropriately reserved to provision the video content blocks to the designated base stations. The moving subscriber can experience no latency in receiving the video content with exceedingly efficient storage usage and reasonable bandwidth expense.

This paper proposes mechanisms to support basic, as well as conditional subscription/notification communication pattern in the Information Centric Networks (ICN), which still leverage and build on the ICN architecture and fundamental design principles with minimum change to the protocols. The proposed new interest message integrates content retrieve request, subscription and notification functions, and adds the conditions field which allow the consumers to specify the notification conditions for their subscribed content. Besides the forwarding face list to the producers for a named content, this paper proposes that the Forwarding Interest Base (FIB) also maintains the forwarding face list to the subscribers and associated notification conditions for the named content. This mechanism maintains the forwarding faces for a retrieval request message and for a notification message of the same named content in the same entry in the FIB. The subscription interest or notification interest does not need to be maintained in the PIT. The proposed mechanism embeds the notification multicast along the forwarding paths, which eliminates the unnecessary notification message delivery by unicast, reduces the notification latency experienced by each subscriber, and improves the network throughput. The conditions associated with each forwarding face to subscribers make sure that a notification message is only forwarded to the face if there were incoming subscription requests from consumers with matching conditions.

This paper observes the fact that an IoT data may be associated with a lot of semantics information. And a consumer usually does not know the name of the IoT data before initializing the request, instead the consumer can specify some major semantics information to be matched by the returned data, e.g. location, time. The paper proposes mechanisms to provide the simultaneous data discovery and retrieval support in ICN architectures. The routing table can be aggregated based on the major semantics information of data. The network resource and consumer's cost can be reduced by giving flexibility to the consumer in tolerating certain degree of semantics deviation between the requested and returned data. We give the analysis on the average number of hops that the interest message traversed in the network with an IoT use case. The performance evaluation shows the significant cost reduction with the increased caching probability in en-route routers as well as the increased semantics information matching probabilities.

In the urban city life, a human usually uses his/her owned user equipment (UE) to request services provided in smart cities. In order for UEs to communicate with the public service providers, the network traffic is generated in the cellular network (e.g. Long Term Evolution (LTE) network) and the Internet very frequently. And the returned results cannot be shared among UEs although they may request the same piece of information. With the information centric architecture applied to the LTE system, the LTE logical nodes can have the name-based routing, the in-network storage and processing capabilities. This paper proposes that by leveraging those capabilities, the LTE logical nodes can be delegated to provide close proximity services to UEs. The paper illustrates the proposed solutions by the use case of restaurant recommendation based on nearby parking availability. The proposed solutions provide collaborations and interactions between different public services in smart cities and significant decreasing in network traffic generated through the Evolved Packet Core (EPC) and the Internet.

Role mining has been extensively used to automatically generate roles for role-based access control. Nevertheless, the two core problems in role mining, role minimization and edge concentration, are both NP-hard. While many approximate algorithms have been developed to solve the problems, experimental tests disclose that no algorithm clearly outperforms the others in both role minimization and edge concentration. The performance results highly depend on the data set under study. To determine the right role mining algorithm, a trial-and-error approach is time consuming due to the computational overhead in mining large data set. We tackle the problem from a fresh angle. Instead of developing fast role mining algorithms, we adopt a data-centric approach that quickly estimates the bounds on optimal role mining results without actually running any role mining algorithm. Based on the inherent features of the data set, the approach can also determine whether it is easy to achieve both role minimization and edge concentration, and if not, which direction, role minimization or edge concentration, that role mining could move toward further.

The recent forecast of billions of devices, all connected to the Internet and generating low-rate monitoring, measurement, or automation data that many end-users/applications frequently request, signifies the need for applying in-network caching techniques to Internet-of-Things (IoT) traffic. Although time delay is not critically important for small-sized IoT content, the expected total traffic load on the Internet from a large number of devices is significant. However, the main challenge as opposed to the typically cached content at content routers, e.g. multimedia files, is that IoT data are transient and therefore require different caching policies. This paper studies in-network caching of IoT data at content routers in the Internet. An IoT data item is uniquely defined not only by its time and location tags, but also a time-range value set by end-users/applications. We provide a model for the trade-off between multihop communication costs and the freshness of a transient data item. Results show that the model can successfully capture the effect of data transiency and can accurately represent the expected gains of a caching system: considerable savings in terms of reduction of network load, especially for highly requested data items.

The built-in caching capability of future Named Data Networking (NDN) promises to enable effective content distribution at a global scale without requiring special infrastructure. The aim of this work is to design efficient caching schemes in NDN to achieve better performance at both the network layer and application layer. With the specific objective of minimizing the inter-ISP (Internet Service Provider) traffic and average access latency, we first formulate the optimization problems for different objectives and then solve them to obtain the optimal replica placement. Then we develop popularity-driven caching schemes which dynamically place the replicas in the caches on the en-route path in a coordination fashion. Simulation results show that the performances of our caching algorithms are much closer to the optimum and outperform the widely used schemes in terms of the inter-ISP traffic and the average number of access hops. Finally, we thoroughly evaluate the impact of several important design issues such as network topology, cache size, access pattern and content popularity on the caching performance and demonstrate that the proposed schemes are effective, stable, scalable and with reasonably light overhead.

Intra-flow network coding (NC) is an innovative technique that has potential to improve multicast performance in wireless mesh networks (WMNs) by allowing intermediate forwarding nodes (FNs) to use coding and overhearing to reduce the number of required transmissions. However the benefits of the NC technology are limited unless there are protocols to exploit its capabilities. The existing intra-flow NC-based multicast protocols are all based upon the conventional end-to-end transport principle. By such a principle, intermediate FNs are unable to accurately determine the minimum number of coded packets they should transmit in order to ensure successful data delivery to the destinations, and hence redundant packets can be injected into the network, leading to performance degradation. Furthermore, the existing protocols cannot handle the bandwidth heterogeneity of multicast receivers very well. We argue that a receiver-driven hop-by-hop transport approach is more suitable for intra-flow NC and these two techniques can create synergy by enabling cooperation among the FNs. In this paper we propose HopCaster, a novel protocol that incorporates intra-flow NC with hop-by-hop transport to achieve high-throughput reliable multicast and to solve the heterogeneous receiver issue. It completely eliminates the need for estimating the number of coded packets to be transmitted by a FN and avoids transmission of redundant packets, as well as simplifies multicast management and congestion control. Moreover, HopCaster employs a cross-layer rate adaptation mechanism that optimizes radio transmission rate in hop-by-hop multicast by taking into consideration next-hop node population changes. Our evaluations show that HopCaster outperforms the existing NC-based reliable multicast protocol.

With the rapid advance in the technology area of data storage, storage capacities have increased substantially while the price has been dropping fast. Motivated by this trend, it has been proposed in the Cache-and-Forward architecture that storage is incorporated into each intermediate CNF router. Content can be cached at CNF routers when they flow through the network, and therefore, routers can serve the subsequent requests later on, without forwarding the requests to the host server, we refer to this caching paradigm as In-Network Caching. In this paper, the content caching is enhanced by Content Broadcast(CB), by which a CNF router broadcasts the information of cached content to its neighboring nodes. In order to solve the problem that with limited storage, how an intermediate CNF router optimally decides which passing content should be cached, we develop a mathematical model for CB to minimize the average content retrieval latency, and propose the Independent Allocation algorithm. We compare the average content retrieval latencies of the proposed caching scheme with two other commonly used cache replacement policies. We study the impact of cache size and locality parameter. The proposed scheme is shown to provide significant performance improvement under various settings by as large as 65%.

With integrated in-network caching diagram, each router advertises the cached content to its immediate neighbors. Although the baseline content broadcast strategy can significantly improve the performance, it leads to low overall cache utilization while each router makes independent caching decisions. In this paper, we enhance the efficiency of content broadcast by providing implicit coordination among neighboring routers. Through detailed simulations, we show that the proposed technique has dramatic performance improvement over the baseline content broadcast scheme. More importantly, this performance gain can be achieved with a minimal communication overhead.

FCC requires that any white space device be able to sense wireless microphone (WM) signals at -114dBm typically corresponding to a SNR of -20dB, an extremely challenging task. This paper presents a novel WM detector based on the ESPRIT algorithm along with an auxiliary prewhitening filter that meets this requirement. Compared with existing detectors, the proposed WM detector not only successfully combats noise, but also provides a technique to mitigate the effect of adjacent channel interference caused by DTV transmission leakage, an issue much less studied.

In order to meet the increasing demands of content dissemination in Internet, we propose a novel architecture for the future Internet called cache-and-forward (CNF), which transports content as "packages" in a hop-by-hop manner towards the destination, instead of transporting a stream of fragmented packets along an established TCP/IP connection. In this paper, we discuss how the CNF network architecture can be designed for efficient content retrieval. We first introduce several specific services provided in CNF network which are centered around content handling and mobile access. We then give an overview of the CNF protocol stack, which is built on top of IP, and consists of a data plane and a control plane. We provide detailed descriptions of each protocol within both planes. Then we present two caching algorithms, where one involves each CNF router making independent decisions on content caching while the other coordinates node caching within an autonomous system (AS) through hashing. Finally, we gave the initial simulation results to show the performance benefits of hop-by-hop transport and content caching.

NDRDL is a software requirements definition language (SRDL), designed as the source language of the software requirements analysis support system NDRASS. It considers both functional and nonfunctional requirements. To avoid the impreciseness and inconsistency of its informal semantics, this paper presents formal semantics of some functional constructs of NDRDL. Different ways and means are adopted according to the inherent characteristics of different constructs.

A framework for a process planning system for the machining of axisymmetric parts using case-based reasoning is proposed. It is composed of four major modules: retriever, modifier, simulator and repairer. This paper focuses on the representation, indexing and retrieving of process planning cases. A feature-based representation for parts and processes is presented. An effective similarity metric is developed to measure the similarity between two parts so that the most similar case can be retrieved rapidly and correctly. It is based on the hierarchical representation of part features. In addition, indexing and retrieving mechanisms are discussed in this paper. A prototype based on the proposed representation, indexing and similarity metric was developed on a Sun workstation using ACIS and 3D-Toolkit from Spatial Technology Inc.