As non-volatile memory (NVM) technologies are expected to replace DRAM in the near future, new challenges have emerged. For example, NVMs have slow and power-consuming writes, and limited write endurance. In addition, NVMs have a data remanence vulnerability, i.e., they retain data for a long time after being powered off. NVM encryption alleviates the vulnerability, but exacerbates the limited endurance by increasing the number of writes to memory. We observe that, in current systems, a large percentage of main memory writes result from data shredding in operating systems, a process of zeroing out physical pages before mapping them to new processes, in order to protect previous processes' data. In this paper, we propose Silent Shredder, which repurposes initialization vectors used in standard counter mode encryption to completely eliminate the data shredding writes. Silent Shredder also speeds up reading shredded cache lines, and hence reduces power consumption and improves overall performance. To evaluate our design, we run three PowerGraph applications and 26 multi-programmed workloads from the SPEC 2006 suite, on a gem5-based full system simulator. Silent Shredder eliminates an average of 48.6% of the writes in the initialization and graph construction phases. It speeds up main memory reads by 3.3 times, and improves the number of instructions per cycle (IPC) by 6.4% on average. Finally, we discuss several use cases, including virtual machines' data isolation and user-level large data initialization, where Silent Shredder can be used effectively at no extra cost.

DNS is an important data source for security for many reasons. If the DNS infrastructure can be brought down, many networking tasks would be impossible to complete. If the integrity of the mapping between domain names and IP addresses is compromised, attackers can redirect users undetectably to IP addresses of their choosing. And malware of many types must in one way or another use the DNS infrastructure as part of their operations. For example, botnets often use fast flux techniques and domain name generation algorithms to rendezvous with command and control servers.

Corporations worldwide work with teams of often dedicated system administrators to maintain, detect and prevent network infringements. This is a highly user-driven process that consumes hundreds (if not thousands) of man hours yearly. User reporting, the basis of most of these incident detection systems suffers from various biases and leads to below-par security measures. In the paper, we provide an approach for near real-time analysis of ongoing events on controlled networks, while requiring no end-user interaction and saving on system administrator's effort. Our proposed solution, ReasONets, a lightweight, distributed system, provides situational awareness in case of network incidents. ReasONets combines aspects of anomaly detection with Case-Based Reasoning (CBR) methodologies to reason about ongoing security events in a network, including their nature, severity and sources. We build a fully running prototype of ReasONets, to demonstrate the accuracy of the system, in doing reasoning and inference on the network status by exploiting events and network features. To the best of our knowledge, ReasONets is the first of its kind system combining detection and classification of network events with realtime reasoning while being capable of scaling up to large network sizes.

The 13 papers in this special issue cover three major areas including privacy enhanced technology, trust and reputation, as well as applications in cloud computing environments.

Network-based intrusion detection systems (NIDS) commonly use pattern languages to identify packets of interest. Similarly, security information and event management (SIEM) systems rely on pattern languages for real-time analysis of security alerts and event logs. Both NIDS and SIEM systems use pattern languages extended from regular expressions. One such extension, the submatch construct, allows the extraction of substrings from a string matching a pattern. Existing solutions for submatch extraction are based on non-deterministic finite automata (NFAs) or recursive backtracking. NFA-based algorithms are time-inefficient. Recursive backtracking algorithms perform poorly on pathological inputs generated by algorithmic complexity attacks. We propose a new approach for submatch extraction that uses ordered binary decision diagrams (OBDDs) to represent and operate pattern matching. Our evaluation using patterns from the Snort HTTP rule set and a commercial SIEM system shows that our approach achieves its ideal performance when patterns are combined. In the best case, our approach is faster than RE2 and PCRE by one to two orders of magnitude.

We provide an approach for real-time analysis of ongoing events in a controlled network. We propose ReasONets, i.e. Reasoning on Networks, a distributed and lightweight system, able to process and reason about anomalies and incidents observed in closed net- works. To the best of our knowledge this is the first system combining detections and classification of network events with real-time reasoning. Our demo will show a running prototype of the ReasONets, demonstrating the power and accuracy of the reasoning process in presence of incidents of various nature.

In this paper we give a new signature algorithm that allows for controlled changes to the signed data. The change operations we study are removal of subdocuments (redaction), pseudonymization, and gradual deidentification of hierarchically structured data. These operations are applicable in a number of practically relevant application scenarios, including the release of previously classified government documents, privacy-aware management of audit-log data, and the release of tables of health records. When applied directly to redaction, our algorithm improves on [18] by reducing significantly the overhead of cryptographic information that has to be stored with the original data.

We describe several new bottom-up approaches to problems in role engineering for Role-Based Access Control (RBAC). The salient problems are all NP-complete, even to approximate, yet we find that in instances that arise in practice these problems can be solved in minutes. We first consider role minimization, the process of finding a smallest collection of roles that can be used to implement a pre-existing user-to-permission relation. We introduce fast graph reductions that allow recovery of the solution from the solution to a problem on a smaller input graph. For our test cases, these reductions either solve the problem, or reduce the problem enough that we find the optimum solution with a (worst-case) exponential method. We introduce lower bounds that are sharp for seven of nine test cases and are within 3.4% on the other two. We introduce and test a new polynomial-time approximation that on average yields 2% more roles than the optimum. We next consider the related problem of minimizing the number of connections between roles and users or permissions, and we develop effective heuristic methods for this problem as well. Finally, we propose methods for several related problems.

Distribution of content, such as music, remains one of the main drivers of P2P development. Subscription-based services are currently receiving a lot of attention from the content industry as a viable business model for P2P content distribution. One of the main problems that such services face is that users may choose to redistribute content outside the community of subscribers, thereby facilitating large-scale piracy. Digital Rights Management (DRM) systems typically employ tamper resistance techniques to control this risk. We propose a system architecture that uses economic incentives instead of tamper resistance to motivate users to keep the content within the subscription community. The key technical contribution we make is to integrate a P2P file sharing service with an escrow service that reliably "pays" the party that is serving up the content. The payment itself can be realized in a number of ways, using "actual" money or bonus points such as frequent flyer miles.Moreover, our architecture facilitates trust between two unacquainted parties by offloading risk to a trusted third party, which can acquire a revenue stream by assuming this risk. To implement the escrow service securely we use cryptographic techniques, such as encryption, hashing, and error correcting codes. Our system motivates users to serve up content of high quality and verifies that users only share legitimate content and not spam, viruses or content that is not part of the subscription. We thereby address other important security concerns in P2P systems and problems like the free-rider phenomenon.

We perform a detailed fixed-point analysis of two-unit recurrent neural networks with sigmoid-shaped transfer functions. Using geometrical arguments in the space of transfer function derivatives, we partition the network state-space into distinct regions corresponding to stability types of the fixed points. Unlike in the previous studies, we do not assume any special form of connectivity pattern between the neurons, and all free parameters are allowed to vary. We also prove that when both neurons have excitatory self-connections and the mutual interaction pattern is the same (i.e., the neurons mutually inhibit or excite themselves), new attractive fixed points are created through the saddle-node bifurcation. Finally, for an N-neuron recurrent network, we give lower bounds on the rate of convergence of attractive periodic points toward the saturation values of neuron activations, as the absolute values of connection weights grow.

This paper presents a comparison study of the computational complexity of the general job shop protocol and the more structured flow line protocol in a flexible manufacturing system. It is shown that the representative problem of finding resource invariants is NP-complete in the case of the job shop, while in the flow line case it admits a closed form solution. The importance of correctly selecting part flow and job routing protocols in flexible manufacturing systems to reduce complexity is thereby conclusively demonstrated.

This paper presents a computationally efficient algorithm for function approximation with piecewise linear sigmoidal nodes. A one hidden layer network is constructed one node at a time using the well-known method of fitting the residual. The task of fitting an individual node is accomplished using a new algorithm that searches for the best fit by solving a sequence of quadratic programming problems. This approach offers significant advantages over derivative-based search algorithms (e.g., backpropagation and its extensions). Unique characteristics of this algorithm include: finite step convergence, a simple stopping criterion, solutions that are independent of initial conditions, good scaling properties and a robust numerical implementation. Empirical results are included to illustrate these characteristics.

In this work, we characterize and contrast the capabilities of the general class of time-delay neural networks (TDNNs) with input delay neural networks (IDNNs), the subclass of TDNNs with delays limited to the inputs. Each class of networks is capable of representing the same set of languages, those embodied by the definite memory machines (DMMs), a subclass of finite-state machines. We demonstrate the close affinity between TDNNs and DMM languages by learning a very large DMM (2048 states) using only a few training examples. Even though both architectures are capable of representing the same class of languages, they have distinguishable learning biases. Intuition suggests that general TDNNs which include delays in hidden layers should perform well, compared to IDNNs, on problems in which the output can be expressed as a function on narrow input windows which repeat in time. On the other hand, these general TDNNs should perform poorly when the input windows are wide, or there is little repetition. We confirm these hypotheses via a set of simulations and statistical analysis.

Recently, fully connected recurrent neural networks have been proven to be computationally rich-at least as powerful as Turing machines. This work focuses on another network which is popular in control applications and has been found to be very effective at learning a variety of problems. These networks are based upon Nonlinear AutoRegressive models with eXogenous Inputs (NARX models), and are therefore called NARX networks. As opposed to other recurrent networks, NARX networks have a limited feedback which comes only from the output neuron rather than from hidden states. They are formalized by y(t)=Psi(u(t-n(u)), ..., u(t-1), u(t), y(t-n(y)), ..., y(t-1)) where u(t) and y(t) represent input and output of the network at time t, n(u) and n(y) are the input and output order, and the function Psi is the mapping performed by a Multilayer Perceptron. We constructively prove that the NARX networks with a finite number of parameters are computationally as strong as fully connected recurrent networks and thus Turing machines. We conclude that in theory one can use the NARX models, rather than conventional recurrent networks without any computational loss even though their feedback is limited. Furthermore, these results raise the issue of what amount of feedback or recurrence is necessary for any network to be Turing equivalent and what restrictions on feedback limit computational power.

Concerns the effect of noise on the performance of feedforward neural nets. We introduce and analyze various methods of injecting synaptic noise into dynamically driven recurrent nets during training. Theoretical results show that applying a controlled amount of noise during training may improve convergence and generalization performance. We analyze the effects of various noise parameters and predict that best overall performance can be achieved by injecting additive noise at each time step. Noise contributes a second-order gradient term to the error function which can be viewed as an anticipatory agent to aid convergence. This term appears to find promising regions of weight space in the beginning stages of training when the training error is large and should improve convergence on error surfaces with local minima. The first-order term is a regularization term that can improve generalization. Specifically, it can encourage internal representations where the state nodes operate in the saturated regions of the sigmoid discriminant function. While this effect can improve performance on automata inference problems with binary inputs and target outputs, it is unclear what effect it will have on other types of problems. To substantiate these predictions, we present simulations on learning the dual parity grammar from temporal strings for all noise models, and present simulations on learning a randomly generated six-state grammar using the predicted best noise model.

It has previously been shown that gradient-descent learning algorithms for recurrent neural networks can perform poorly on tasks that involve long-term dependencies, i.e. those problems for which the desired output depends on inputs presented at times far in the past. We show that the long-term dependencies problem is lessened for a class of architectures called nonlinear autoregressive models with exogenous (NARX) recurrent neural networks, which have powerful representational capabilities. We have previously reported that gradient descent learning can be more effective in NARX networks than in recurrent neural network architectures that have "hidden states" on problems including grammatical inference and nonlinear system identification. Typically, the network converges much faster and generalizes better than other networks. The results in this paper are consistent with this phenomenon. We present some experimental results which show that NARX networks can often retain information for two to three times as long as conventional recurrent neural networks. We show that although NARX networks do not circumvent the problem of long-term dependencies, they can greatly improve performance on long-term dependency problems. We also describe in detail some of the assumptions regarding what it means to latch information robustly and suggest possible ways to loosen these assumptions.

Many different discrete-time recurrent neural network architec(cid:173) tures have been proposed. However, there has been virtually no effort to compare these arch:tectures experimentally. In this paper we review and categorize many of these architectures and compare how they perform on various classes of simple problems including grammatical inference and nonlinear system identification.

We introduce and study methods of inserting synaptic noise into dynamically-driven recurrent neural networks and show that ap(cid:173) plying a controlled amount of noise during training may improve convergence and generalization. In addition, we analyze the effects of each noise parameter (additive vs. multiplicative, cumulative vs. non-cumulative, per time step vs. per string) and predict that best overall performance can be achieved by injecting additive noise at each time step. Extensive simulations on learning the dual parity grammar from temporal strings substantiate these predictions.

The TNM staging system has been used since the early 1960's to predict breast cancer patient outcome. In an attempt to in(cid:173) crease prognostic accuracy, many putative prognostic factors have been identified. Because the TNM stage model can not accom(cid:173) modate these new factors, the proliferation of factors in breast cancer has lead to clinical confusion. What is required is a new computerized prognostic system that can test putative prognostic factors and integrate the predictive factors with the TNM vari(cid:173) ables in order to increase prognostic accuracy. Using the area un(cid:173) der the curve of the receiver operating characteristic, we compare the accuracy of the following predictive models in terms of five year breast cancer-specific survival: pTNM staging system, princi(cid:173) pal component analysis, classification and regression trees, logistic regression, cascade correlation neural network, conjugate gradient descent neural, probabilistic neural network, and backpropagation neural network. Several statistical models are significantly more ac-

Characteristics of error surfaces for the multilayer perceptron neural network that help explain why learning techniques that use hill-climbing methods are so slow in these networks and also provide insights into techniques to speed learning are examined. First, the surface has a stair-step appearance with many very flat and very steep regions. When the number of training samples is small there is often a one-to-one correspondence between individual training samples and the steps on the surface. As the number of samples increases, the surface becomes smoother. In addition the surface has flat regions that extend to infinity in all directions, making it dangerous to apply learning algorithms that perform line searches. The magnitude of the gradients on the surface strongly supports the need for floating-point representations during learning. The consequences of various weight initialization techniques are also discussed.>

P-type nodes are capable of computing the exact a posteriori probability for the multiclass Gaussian problem. A theory that extends to problems where some classes are given by a sum of multiple Gaussian kernels is presented. The corresponding weight solution can actually be achieved through learning since this weight solution is a minima of the mean squared error criterion function. A special type of problem, called the 1/ΣMproblem, which demonstrates the capabilities of P-type nodes is presented. The authors show that the P-type node can solve a number of interesting problems, including theI-4i-16Iproblem, the XOR problem, a multimodal non-Gaussian problem, and a one-class classifier problem